Stability is unlikely
It is exceedingly unlikely the balance of power around "mana recharge time" would stay at the same stable point after an industrial revolution scale change in how mana is recharged.
Any mechanism would have to be extremely contrived.
It would be like the balance of power between ranged and melee combat being maintained after people invent the AK-47 (or even a more primitive firearm).
This doesn't mean that the system will become unstable; the system will find a new stability. The transition period will often not look very nice.
Warfare still exists post-AK-47, it just doesn't look at all like pre-modern warfare. And people still own melee weapons. For the longest time they even mounted them on their ranged weapons. But the pre-gun balance between melee and ranged combat is completely gone.
Now, in that bloody transition, a form of stability could occur, where both sides still have weight. Inertia is a strong thing.
Real-world example
We can look at the industrial revolution itself. Prior to it in Europe, the primary wealth production was by a plague-ravaged depopulated populations with subsistence peasants whose surplus was harvested by the ruling classes. On this surplus the rest of the economy floated.
Stability was produced by the Malthusian trap, where more peasants increased production but produced less food and surplus food per capita.
The plague depopulated lands had a higher per capita surplus, allowing greater non-farm worker populations, and efficiencies of various non-farm productions grew (the practice effect, basically).
Land owners, able to produce almost as much food with fewer peasants (or other resources; highland grazing of sheep, for example), and with demand for production in urban areas, would kick the peasants off their land for them to migrate to the cities and join the industrial efforts and/or starve.
Over time this did lead to increased agricultural productivity per unit land and the increasing urban industrial productivity led to higher standards of living, but in the short term it was a bloodbath.
Fictional world effects
We'd expect a massive shift towards the economic power of class C witches, which often results in a shift in political power. Such a shift is resisted, leading to a bloodbath of violence. Either class A surrenders, or class A sees this coming and contains class C before it is too strong to stop.
This may not play out as class warfare, but as a proxy; sub-societies that that values C over A or A over C will have their relative power shift pretty rapidly.
As an example, suppose you have a country where all 3 classes have equal power, and a backwater where C has more power for legacy reasons. In the before-revolution, the first country is 3 times more powerful than the 2nd, and the 2nd only holds on because of geographic advantages of its position (defensive ones), or is a client state.
Then this revolution occurs, and the economic strength of the 2nd nation starts growing year after year. At some point is starts to enter into the same "class" as the first nation.
The first nation will have placed restrictions on the 2nd nations powers (colonies, resources, whatever). The 2nd nation will chafe at these, and start violating those restrictions. The first will respond. And the conflict will escalate.
Typically it comes to a head with violence; the first nation will draw a line in the sand, the 2nd will cross it, the first will back down and draw another. Eventually the first nation will hold its ground and conflict will result. If this happens early enough, the first nation will win (at political cost). If it happens too late, the second nation will win (and treat its former oppressors impolitely).
If the 2nd wins, we have a new dominant power, and the world switches over to its economic model.
If the 1st wins, we instead get a period of brushfire oppression, as other areas start mimicing 2nd nations economic model and out pacing the 1st nation's "balanced" approach. It then has to go in and smack it down.
Odds are a 3rd party nation, with a "balanced" approach, will be less zealous about maintaining its "balance" between the 3 classes of witches. The 1st nations adventures in oppressing class C-dominant witch societies will drain it, and the 3rd nation's relaxed approach to its society leaning towards class-C dominant will strengthen it.
Eventually the 1st nation falls apart from its failing economic performance, or there is a war between the 1st and 3rd nations over the 1st nation's over stretch, and it collapses.
(Alternatively, the 1st nation ends up stealing the 2nd nation's class-C dominant economic model over time, "becoming what they fought to defeat").
Pseudo-stable possibility
Now, in a story, you could maintain "balance" in a period after the first war above.
The 1st nation was the dominant one prior to (or in the initial parts of) the technological revolution, where all classes of witches are equal. Call this one Empire.
The 2nd nation was a "dark power" whose "cabal" tried to take over the world with "dark magic" fueled by mana potions. In this nation, class C witches are dominant (Propaganda or not, this is what the 1st nation would claim in order to encourage sacrifice). Call this one Upstart.
Empire wins the war over Upstart, oppressing the "class C first" dark magic philosophy of Upstart. Since then, they are fighting "dark witch" infestations world wide, and even domestically, as people start using the "dark magic" fueled by mana potions to generate more economic power (cast by class C witches).
Empire has no problem with the "grey magic" that class C witches can cast; this is the traditional balance, where class C witches cast different kinds of spells than class A witches. That grey magic is the things that class A witches cannot (who cast what I'll call "white magic"). It is only when the class C witches use mana potions to replicate class A witch magic -- what the nation calls dark magic -- that they step in and stop it.
What they call it exactly doesn't matter. In order for you to have a pseudo-stable situation, you need a force suppressing this new, economically powerful strategy of replacing class A witches with class C witches + mana potions. As there was a war against an Upstart regime that used this tactic, they will have villianized it somehow (to justify the costs of the war), and continue after the war.
This results in a social structure backed by violent oppression that maintains the monopoly of class A witches on a certain kind of magic. By blocking class C and B witches from doing that kind of magic, we maintain that class A witches have unique economic value. This preserves the balance between A and C (and hence B) witches.
The class A witches are the majority of the dominant countries population, and control a lot of the economy and politics. So this state-enforced monopoly is popular, and justified by claiming that the "dark magic" is evil and wrong.
The oppressed "dark magic" witches (in country) are forced to act like a criminal organization; sort of like drug dealers, if you are divorced from the enforcement mechanisms of the state, you have to do your own brutal ones. Every such "organized dark magic" crime is more evidence that the state is right in killing/imprisoning every dark magic witch.
(Any powerful economic activity, like class C witches replacing A with cheap mana potions, is going to generate a black market. If it is officially illegal, those doing it will form a criminal organization in order to enforce rules and protect themselves from the state. The worse the oppression, the bloodier the criminal organization.)
Overseas, Empire will support nations oppressing dark magic, and overthrow nations that are soft on it; in many places, the penalty for even reading about it is death or worse. If they don't, dark magic economies will grow and overthrow Empire; they have to be burned out before they grow.
Meanwhile, experiments in dark magic are done in controlled fashions, or hidden, at various institutions within the Empire; dark magic is going to be economically powerful (hence useful) and probably able to do things that conventional magic cannot.
A rival similar-scale nation to Empire is probably going to experiment more with dark magic than Empire is. Rumors about it will swirl, and the villianization will raise tensions. (This is one of the reasons why this isn't a stable situation, even with really brutal oppression; but you can pull it off for a medium long period of time).
And meanwhile, the bounds of "grey magic" leak into dark as the state uses it to maintain its power (against internal conspiracy and external threats).
Characters in your fiction, even the "dark witches", might take this framework as a given; that "dark magic" is evil. Despite there being noting "inherently" evil about dark magic. Many of its practitioners will be doing horrible things in order to form the required social bonds while being treated like criminals, which in turn is used to justify the oppression, which feeds back on itself.
TL;DR: A Solution
The main power of your world engages in violent oppression of class C witches who infringe up on a monopoly granted to class A (and B) witches. This violation was only economically feasible after mana potions existed.
This monopoly maintains the economic balance between A B and C.
Those that openly violate this monopoly are villianized and delt with extremely hashly. Nations that don't maintain this monopoly are invaded, overthrown, and (as a last resort) destroyed in righteous fury.
All powers actually violate this monopoly, because the economic advantages are large enough, and there are going to be strategic applications of mana-potion fueled class C magic that no state can afford to ignore. Each top-tier power has rumours (believed by their enemies) that they are engaging in the oppressed form of magic. Meanwhile, "edge cases" are moved from unacceptable to acceptable in some areas, backlashes occur, and blood flows everywhere.
Presuming mana potions continue becoming more efficient, the pressure to contain the new advantages class C witches have will eventually grow to the point that entire social structures crumble. But in the meantime, you can have a pseudo-stable situation.