I'm not impressed by much in the virtualization space... they're all done with their two valid use cases and are trying to expand into unnatural places. The valid use cases are the obvious data center centralization and what I'll call general geekery. Data center centralization is a backlash against the common sysadmin practice of one app per server; a valid operational move that wastes server resources, which are neatly recovered by going to virtual servers. General geekery is a broader topic... developers, sys-admins, services people, sales engineers, QA people, security testers... there are a lot of people who need virtual machine technology on their desktop, and their needs are nicely taken care of with the available tools.
So, where to expand? How to sell more? The juiciest remaining space is the enterprise desktop, so let's target that. There's a lot of variations on thin-clients accessing a locked down data center... well, the people whose needs are met by that use case already have tools in Citrix and Terminal Server, so you've basically got a slight broadening of that market and a 'brown field' incumbent replacement strategy. Then there's the local virtualization ideas, PCHV type stuff... in which the argument is that OSes have failed to adequately separate applications, so we should make things a lot more complicated so that our applications will really be separated. I don't see how this is supposed to make things easier, cheaper, or shinier, and those are the three reasons customers buy things.