i agree with grazr about environments. most maps here are being developed for pub play and it makes no sense to test a map that is supposed to end up being played with wranglers, sandviches and whatever the hell else in a vanilla environment. crits and damagespread only skew the battle outcome results a little bit, different weaponry changes the way battles play out.
That was the same argument people here used in defense of crits/spread, and it holds just as little water as it did before. Now considering the wrangler is blatantly overpowered in just about every situation anyway, there isn't a whole lot map authors can do to make it "less op" (same goes for the sandvich and just about everything else), and testing it in that environment is more likely to throw the results of the test away from what they otherwise would be, and allowing it causes more grief than good. The fact of the matter is that the unlocks that stay the least bit true to the original game design have very little (if any) effect on the outcome of a map.
As I said before, there is very little the mapper can do to handle obviously broken unlocks, such as the wrangler, sandvich, etc, and the responsibility to fix them lies
only with Valve, and no one else.
I find the whole, "but that's the environment they will be tested in", argument to be ridiculous, since first off, that environment changes every week anyway by adding more weapons that can have ridiculous effects on gameplay. Consistency in tests goes out the window for any map tested over a period of more than a week (this is a bad thing), making it even harder to get accurate information from a series of tests. The only environment that remains consistent is one that just outright bans any new shit straight off. If you don't think so, please try to seriously argue to me that maps like granary, badwater, etc, which were not tested in the current environment, are now trash maps. Good unlocks do not radically change whether good map design is still good map design, bad unlocks on the other hand do, and there is fuck all map authors can do about them.
The best possible testing environment is one that returns consistent results, since that's the only way you're going to be able to tell whether something's wrong with your map, and not just a shittily designed unlock throwing it off.