Yes. According to this post on the comp.std.c++ newsgroup and WG21 meeting notes the proposal on a new allocator model has been rejected.
I've seen a talk on this model on the C++ Connections conference in Las Vegas. Back then in November I didn't like the model because it didn't work with copy-constructors, (N)RVO and yet to be added move-contructors. However Bjarne Stroustrup was a strong supporter of it. That got me worried. But I assumed that the committee would do a right thing (either make it work or punt it). And it did.
