Creatures would have to either live under deep canyons that have been eroded by time,
Well, what is doing the erosion? Wind and sand? Probably not very good places to shelter!
or be able to actually dig there way into the glass.
You've already stated the glass is "super-hard". This rather limits the ability of any creature to dig into it. Even if they had evolved harder-than-glass claws (corundum, maybe?) there's little point to doing so... they'd be better just leaving such a hostile environment instead of wasting all their energy digging somewhere that has no food and no water. If there is food and water, it'd have to be surface water (hard to erode glass, and it isn't very porous) and anything living there would be in the water or on its surface. Probably just algal gloop, in the absense of a good supply of nutrients from outside the glass region. Don't need claws to eat gloop!
Superheated creatures that melt the glass and sand (maybe even eating it by breaking it down into an edible form)
Melting glass takes a lot of heat, and a lot of energy. Where is this energy coming from? You can't realistically break down glass into an edible form as it has already been quite thoroughly oxidised. If you had access to hydrofluoric acid and an HF-based biochemistry you'd be able to oxidise the glass, but you'd need a ready supply of rare fluorine-based compounds, and you're unlikely to find a lot of them just lying around in a big plain of regular glass. They'd effectively just suffocate, surrounded by bland, low-nutrition food. Doesn't sound very appealing.
Literally swimming through channels of light focusing glass.
Leaving aside the possibility of natural fibre optic channels forming in natural glass, signals don't "swim" down optical fibres. They can't decide to change direction or turn around. They just propagate down the channel to the end where they are emitted or absorbed or reflected, depending on what is there. Signals don't last forever, either... there's always some loss and absorbtion, even in the highest quality optic fibres. Very long fibres have devices spaced along their length to "regenerate" signals.
That said though, there is some scope for something weird to live in there. Have a think about the nature of Stephen Baxter's Qax, and maybe phonons in the glass that propagate at merely the speed of sound in glass, not the speed of light...