It's a promising project...for trying to support the Windows OS directly on Linux a' la OS2's support for the same.Today I stumbled over a short report on Wine vs LUK benchmarks. LUK, or Longene, is a Linux kernel module implementation of the NT system calls (rather than intercepting them in userspace as Wine does). The speed differences seem rather dramatic!
It would be interesting to get some independent benchmarks on native Windows, LUK and Wine using Phoronix.
As a technology demo, it's spectacular. As a production answer for things like your games, probably not so much so. Currently with the 0.3 builds, the instructions for install tell you to do things like turn SMP off. Ouch.
I'd say this would be fairly counterproductive just to make WINE honest about it's naming (WINE is/isn't an emulator. It's misnamed. WINE the library isn't an emulator- it's an abstraction library for the Windows API, just like they claim. WINE the application is a program loader that emulates the Windows loader behavior and dispatches calls to the wineserver to make it look like you're running under Windows... With LUK, you merely (at this time, mind...) get rid of the wineserver and the bulk of the loader emulation pieces, making it "not an emulator" again.)- with the gains you'll get via SMP, etc. on the stock WINE side of things, most machines will be ahead of LUK except for things like this little benchmark you linked to- at least for right now.
I think we, as the community, ought to watch this and see if they get anywhere with it- but I don't think it's a good use of anyone's time until they can get a bit further along on things with it. I'd say with their rate of development, if it continues as it has, they'll be ready for something that would be worthy of more than the occasional look-see to see where they're going with it.