Meaning using 3 min charts makes you profitable but using 2 min charts makes you unprofitable. Would this mean you have a crappy system, or a very fine-tuned system?
No, it is not robust. And in case, you don't know what that implies, here is one person's idea: http://www.michaelcovel.com/archives/000021.html
2 or 3 min time frame makes no difference for a good system. 1 and 5 min does. Signals will be quite different
Two min is too choppy. Three min is just right. I should try four. Ok, keep in mind this is for my latest "system" that I came up with. I'm sure some backtesting will render it useless anyway.
if trading on that small a time fraim, you should be using tick data and streaming prices to test/trade, not chart signals. also paying almost zero commisions.
Obviously, you are talking about daytrading and in that case you are looking at only one piece of the puzzle. There are at least three more parts, that I know of, which are more important.
2 min vs 3 min may produce different signals based on how you are trading, but I would guess that the difference in performance would not be terribly different.
I can't even imagine a signal that would work on a two minute chart that would fail on a three minute chart. If it fails that quickly I would presume it wasn't much of a system. If a signal works it should work under all time frames.