I think that statement may be a bit too strong. There are some major benefits (both in perf and programmability) to applications using MSAA. It's very difficult to gauge what the future adoption rate will be. 10.1 came with SP1, so hasn't been available to users for long at all. This combined with the relatively small number of D3D10-exploitive titles (as opposed to D3D9 ports) up to now means that most end-users aren't yet aware of these improvements.
Assuming DX11 is a superset of 10.1, and that NVidia is likely to make a DX11 part, one could say that NVidia will make a "conforming" part, but it's just a matter of how late they are to the party. It's all a matter of timing, and depends on adoption rates of the 10 API, Vista, and their development schedules.
