Fundamentally I think that what we disagree upon is the approach. If I'm hearing you correctly, I say zero tolerance, you say legitimize but compartmentalize.
The difference is that if the site retains the zero tolerance course while the detection methodologies mature in order to get a better handle on the feasibility of actually stopping it the option to legitimize and compartmentalize down the road will still be available if it turns out that it can't be stopped -- which I do not believe to be the case, incidentally.
If, on the other hand we legitimize but isolate now, well, that's a balloon that just can't be un-popped.
I think that Erik more than deserves our patience.
If I report an user I am 100% sure he cheats, I expect him to be banned in one-two weeks. If this does not happens, it means that there is a clear discrepancy between what I think is cheating and what chess.com thinks it is cheating. If the discrepancy is so big, then clearly there is no point for me (and others) to report suspected cheating.
I reported someone more than a month ago. Back then he was like 1500, with 30-40 games played. Now that person is 2100+, with 400 games played. Really, after that person cheated almost every player in live chess, there is no purpose to ban him now. If chess.com wanted to close his account, they should have done it more than a month ago, before he inflicted so much damage.
By the way Grobe, you say that I am a PR nightmare. What about the cheaters? Let's say Victormpm, who cheated in THOUSANDS of games. I am a nightmare, the cheaters are beneficial for the site, isn't it? :))