Yeah... those plates need to be well-submerged...

There was a time when boiling off water was an expected thing.  That's when generators used mechanical regulators, and running long hours doing field work at goverened speeds with your lights off would boil them down.  Plate metallurgy was a little different then, and they made thicker plates and heavier busbars back then... but nowdays, they're thinner, and more dependant upon good regulation.  If you have an electronic regulator and the wiring is a little weak somewhere, the regulator will get an improper feedback signal (voltage looks low) so it will increase the field current to force output voltage up accordingly... and you'll wind up boiling 'em.

The really, really serious thing to watch out for, is when the cells' levels are INCONSISTENT... if you have low cells, and high ones, the high cells are frequently partially shorted (sulfation, contamination, freezing etc), thus, they're not charging, and the remaining cells are thus being subjected to a proportionately higher charging voltage and current.

An aside (and somewhat relevant) note... if there's a possibility that you've ever had the battery freeze, remove it from service.  Not only can freezing short plates of a cell, they can also fracture the inter-plate busbars... and when this happens, the busbars will exhibit high resistance, or even arc at the fracture, which is inside the battery casing, up where all the hydrogen gas hangs out...  just a little oxygen in there, and you have an explosion.

So in February both batteries charged right up in a couple hours. Then in March the back one, the one that took so much water, wouldn't take a charge after a day and a half. So I took the tender off it, and load tested it. It went from weak to dead. So I took Curt's advice and put in a big battery. I'll use the other battery on my tractor diesel fuel tank pump. Thanks Curt, and you other guys too.

 

 IF levels in cells keeps getting low was always told it's a sign of overcharging or a cell is starting to short. 

Yeah, a battery's electrolyte distribution tells you about the health of the battery.

If the electrolyte in all cells is even, that means the cells are all working together.

A cell that looses electrolyte rapidly signals that the others (which aren't) are not doing well, thus, charging current is being carried through, and applying a higher charging load to the one cell that's consuming more water.  If the cells all tend to maintain a proper level, A hydrometer will indicate the electrolyte's specific gravity, and in that, it's cell state-of-charge.  They should all be about the same.

Common cause for bad cell, is a shorted plate, contaminated electrolyte, or fractured busbar.  Internal physical damage is most often the result of freezing, but there's others.

If you suspect the battery has ANY sort of problem, remove it from the machine, and treat it like a bomb...  They tend to explode when one attempts to charge them.