I guess his original point was that while moving to 64 bit would help the few users who actually hit memory allocation limits, it would also be a monumental eff

I guess his original point was that while moving to 64 bit would help the few users who actually hit memory allocation limits, it would also be a monumental effort for a project the size of Visual Studio to do so, including out-of-process support for 32-bit extensions to maintain compatibility, probably rethinking designers and debugging along the way. And all that to make things actually slower and taking more memory for everyone. His argument wasn't entirely »Stop using those tools while we do nothing at all«. It was »They should have done the right thing, but we'll also work towards making Visual Studio take less memory«. Especially considering that not leading everything you might eventually ever need at once in the beginning has probably a much smaller compatibility impact than upgrading significant parts of the IDE.