The drawbacks
Your drawbacks are somewhat flawed or incomplete. Let me elaborate:
- If the habitat is meant to move, the momentum imparted to the water needs to be accounted for when it comes time to apply the brakes. Rendezvous With Rama "solves" this with a sheer cliff on one side of the sea, against which the water sloshes. This cliff towers over the rest of the habitat and serves no other apparent purpose.
The mass of the water is a valid argument. Water is rather heavy, and takes a lot of energy to acc/decelerate. However, note my feedback in 4.
I'm not sure how this cliff is intended to be used. Sloshing against a cliff will not help with the braking problem as far as I'm aware. Water sloshing back and forth is a zero sum game, energy-wise. And during a braking action, any force that the cliff exerts on the water (to slow it down) means that the water's kinetic energy is being transferred to the cliff (action equals reaction, thus making it a zero sum game).
The cliff does function as a containment vessel for the water, but that's not related to the kinetic energy argument for space travel. Assuming that the space habitat is able to brake slowly for a long time, the inertia of the water may even be negligible enough to not need high cliffs (but you would need to brake really slowly to achieve that)
- The sea takes up space – nearly 20% in Rama's case – that could be dedicated to habitation, agriculture, and so on.
The benefit of having the sea should be weighed against the benefit of what else to do with the space. You wouldn't buy a house with a massive garage and a tiny kitchen, if you cook elaborate meals every night but do not have a car.
If you so choose, it's also not impossible to make the bodies of water subterranean (i.e. build your agriculture on top of it), unless you need open access (or line of sight) to the water.
- The sea, especially if it occupies the circumference of the inner surface, is a barrier to all surface-based transportation.
Am I to believe that we're capable of building a generation colony ship slash space habitat, but shudder at the thought of having to build a bridge?
Any reasonable surface-based transportation is going to condense traffic on roads (instead of driving wherever you want). It's not excessive to build a few bridges.
Keep in mind that many of the logistical constraint of transportation on Earth is related to the transportation network having evolved and expanded. Many cities (especially in Europe) were built with a horse and carriage in mind, not a truck with a container on it. It's impossible to move the buildings around to widen every road; and this is the main issue with the increased traffic load on roads.
However, the space habitat will be built at a time where everyone is already aware of transportation. Roads can be built before residential areas; and there will be no logistical issues unless the space habitat is expected to undergo major redevelopment while being habited at the same time.
A great example of this in real life can be found by comparing the US city layout to a European city layout. When most US cities were built, humans already knew the concept of traffic much better than when European cities were founded. The road system reflects that knowledge, as gridsgenerally handle heavy traffic better than Europe's snaky roads.
- If the primary purpose is water storage or recreation, then arguably numerous smaller and less obstructive lakes could do the same job with fewer or reduced downsides. In fact, smaller lakes could be even more effective (e.g.: if the goal is recreational coastline, ten lakes 1/10th the size of the sea would have more than three times the sea's coastline; water stored within the hull does not pose a flood risk).
If the primary purpose is water storage, then that means the water is needed. That means that even if there was no body of water; that the water would still have to be on board somehow (e.g. in small containers, all over the place).
If the water is on board anyway, and we're only considering whether we store it in a large body or keep it distributed, then the kinetic energy argument (see 1.) is moot.
The same amount of water will have the same mass, regardless of where on the ship it's located.
Looking at recreation; most of the reason why humans enjoy water-based activities is because it simulates a lower-gravity environment (due to our bodies' buoyancy). But in the space habitat, we're already in space. It would be possible to create a (near-)zero-G zone near the axis of rotation of the ship; thus creating a surrogate for water-based recreation (and arguably even more impressive by human standards, at least initially before it is considered normal).
Taking this into consideration, recreation doesn't seem like a good enough reason (by itself) to warrant bringing an additional body of water.
- Water is less dense than the surface and sub-surface material it replaces, which could throw off the balance of the habitat and cause it to spin end-over-end.
Water is less dense than steel (or any similar metal alloy), but keep in mind that you don't need that much steel. Steel has a very high structural integrity, and the mass of that construction can be dramatically reduced by using bracing and crossbeams, as opposed to simply using solid steel.
This is something that humans have mastered for milennia. The Ancient Egyptians already understood the benefit of a wheel with spokes (compared to a solid disk), and they weren't even the first ones to realize that.
which could throw off the balance of the habitat
You're putting the cart before the horse. If the space habitat is going to contain bodies of water, that will be known at the time of building the space habitat. Therefore, the blueprints can account for the water.
It also doesn't make sense that you'd complain about the uneven distribution of water, while ignoring the presumably equally uneven distribution of housing (unless you want to house people based on body weight and constantly reshuffle them?), agriculture or people moving around. A couple of people moving around are insignificant, but what about a massive festival with people jumping up and down?
Secondly, the only balance that is needed (long-term) is an even distribution around the axis of rotation. Going by the O'Neill cylinder, all you need to do is ensure that every "arm" of the habitat contains roughly the same amount of mass.
In other words, if you give every "arm" a body of water of the same size (= mass), the balance is maintained, regardless of how big/small the lakes are.
Summary
- Any drawback related to the space habitat not being able to house the bodies of water, are negated by the fact that the space habitat will likely be built with the bodies of water in mind. You don't build a ship, only to then decide what to put on it.
- This includes any issues with navigating around a body of water. The roads can be built before the water is added to the ship (here on Earth, the water was there before the roads were built). This gives us the benefit of designing the roads and lakes in a way that does not compromise logistical efficiency.
- The kinetic energy argument is valid. However, this relates to the total amount of water that is on board, regardless of whether it's condensed into large bodies of water or comparatively small water tanks. Mass is always expensive in space travel. If a space habitat is being built, everything we would want to take with us would have to have its benefits weighed against its mass (pun intended). E.g. we're not going to take statues of lead or gold, just because they're pretty. The drawback of the added mass far outweighs the value of bringing it with us. In other words: regardless of having large bodies of water or not, we are already inherently constrained in the amount of water that we can bring.
- Non-functional uses of water (recreation) are generally negated. We're already capable of having a zero-G (or low-G) environment in space, which is very similar to floating in water. Also, if the people are only willing to travel around space if they're able to go jetskiing every weekend; they're bringing the wrong mindset to an interstellar journey. It is more than reasonable to ignore any request for facilitating water-based recreation.