Water bath filling is a controversial subject.
Believe me when I tell 'ya that tanks get hot when filled. Even if filled slow. For example, my HP100s, filled by my 3.5cfm compressor (not "fast" by any means!) get quite warm during the fill process. I have "shot" them with my IR themometer at their final pressure and have seen temperatures - with a fan blowing on them the entire time - of right around 100-110F, with an outside air temperature of 80F.
My AL80s are "cooler" outside than the HP steels. But this is misleading; steel is a better thermal conductor than aluminum, and the walls are thicker on the AL tank! Thus, the GAS in the AL tank is probably at the same (or even higher!) temperature than the gas in the steel (since some of the energy bleeds off through the walls in the form of heat in the steel tank more quickly!) This has been confirmed by filling both to the computed "should be" final pressures, accounting for the heating, allowing them to cool to room temperature, checking what THAT should be at current ambient, and putting a (digital) gauge on them - they come out "spot on."
This is with ROOM TEMPERATURE air post-filter-stacks - the fittings and hose are at AMBIENT (verified again by IR thermometer) - so its not the gas going IN that is hot, its the compression of the gas in the tank that makes it hot. If you fill off a bank, and do it fast, then the problem is even worse. Its pretty easy to get a REAL hot fill by slamming in the air from a bank; that's a bad practice.
The use of water baths is an attempt to prevent the heat, but it does nothing about the actual issue, which is that you're filling the tank too darn fast!
There are, however, problems with water bath filling that go beyond the issue of fast fills. The most serious of them is that if you get even a TINY bit of water in the fill adapter, and then put it on a tank, you have just blown that water DIRECTLY into the tank. Since the whips are right there AT (or even IN, if the whole thing is submerged while filling!) the water, this is a very, very real possibility. In fact, I suspect its a near certainty over time that this will happen to a tank filled in water.
My personal view (and that of PSI now days) is that you should fill DRY - NOT in the water - and if the tank is uncomfortably hot you're filling it too darn fast! SLOW DOWN! Of course that means that a shop has to take longer filling tanks, which is contrary to getting the most fills done in the least time.....
Second, the rated pressure is at 70F. If you are going to dive in cold water, you have the same cubes of air, but less pressure. This is simply the way the gas laws work. Similarly, if you have HOT tanks (say, 90F) and water, then your tanks should be under MORE pressure but still have the same capacity in real cubic feet (at 70F) terms. Tank ratings take this into account; it is thus "safe" to account for it while filling, PROVIDED the tank is not outside its service temperature limits (typically 120F is the "ceiling" on such.) So, for a tank at 120F, the pressure should be roughly 9.5% above "rated"; if you have a tank rated for 3000 psi, and at "final fill" time its at 120F (maximum "normal" temperature), then the final pressure at the fill station should be 3285.
When it cools to 70F, it will read 3000 psi.
(The salient gas law issue here is that pressure in a fixed-volume vessel is directly proportional to the temperature of the gas in degrees Kelvin. Kelvin is Celsius + 273; that is, 0C = 273K. While scuba tanks are not QUITE "constant volume" vessels - they DO stretch a very small amount when filled, for all intents and purposes it is safe to treat them as "constant volume" containers.)