Because of these limitations, I am almost tempted to mutter the old cliche that the amazing thing about a dancing bear is not that it dances well, but that it dances at all. However, that would be unfair. While addressing these limitations should make the final version of UVD more polished, most of them are minor enough to make little difference. For the most part, UVD performs superbly and, for old GNU/Linux hands, provides the mildly subversive thrill of seeing KDE apps running under Windows as well as native ones. You simply aren't supposed to be able to do things like UVD.
The only real question that UVD leaves me with is who such efforts are for. You can understand why developers would appreciate the challenge, but are that many Windows users actually clambering for an easy way to try GNU/Linux? Live CDs for GNU/Linux have been widely available for several years now, and they have yet to create a mass exodus from Windows. Perhaps, though, their relative slowness compared to an operating system running from a hard drive makes them less impressive.
Another worry: Given the success with which UVD integrates with the desktop, any curious Windows users who do exist might not fully appreciate what they are seeing. Equally, they will miss seeing features like virtual workspaces or the number of customization options that distinguish GNU/Linux.
Perhaps the real audience for UVD is free software users who are forced by work requirements to use Windows, but would rather use their favorite applications. While the lack of GNOME applications or, in some software categories, multiple choices might limit this market, it could be a substantial one.
None of which makes Ulteo Virtual Desktop any less ingenious. Ulteo may need to worry about its audience so they can monetize it, but the rest of us can simply appreciate the programming and be grateful for the addition to our toolboxes.