I think this 3 week deadline was possibly the last (or at least the penultimate deadline). If core are still flaky, or set a timescale too far out I think it will be over within a week. If core say "we'll do a HF in <a month or so>" the chines may accept it, if its not too far out. If core subsequently don't deliver on that I think that is when you will see the miners settle on a non-core HF. Classic or otherwise.
I also think that core is playing a very dangerous game, because if the Chinese are pushed that hard, then they *will* do it in house. Ask yourself, wouldn't you?
Sorry to sound like a broken record. The maximum Core will do is lie, saying an HF happens after SegWit.
Core has driven into a dead-end by denying the ecosystem a simple block-size increase, for which a dozen BCT and reddit polls showed >80% support. A lot of those voices were non-technical, but the issue is basic and everyone can understand it. So the community consensus was (and is) clear. Sure a few hold-outs did not like this proposed change, just as some did not want soft-fork P2SH, or LevelDB, but got them anyway.
By delaying Core made a fatal error*, it upset a large part of the community and accelerated the take-up of alternative implementations. Every week the devs delay hard-forking they entrench opinion against them and make it less likely that Core would come out of a HF with anything other than a minority or rump implementation.
They are not playing a game, they are in a trap of their own making.
* (Maxwell almost certainly deserves most blame for their predicament).