[TamiW]

The plural of anecdote isn't data. Even the polls, at most can say people with or without skimmers have or have not had occurrences of gas bubble disease. It would be great for designing a controlled experiment, and that is pretty much it.

The problem with the skimmer-cause-hypothesis is that it discounts the cases where adding a skimmer has solved gbd. You cited public aquariums having trouble until they removed skimmers above, but I know of a few I talked to the staff at directly that had trouble until they added skimmers (but in many cases it was part of a system redesign, so who knows).

I think you're right though, it's a complex problem and it's likely to not have a single answer. We don't even know yet what the gas is that is causing the problem. CO2 is the obvious choice considering the effectiveness of diamox. However, O2 would make more sense if it was due to a skimmer, and N2 if its due to supersaturation from cavitation.