Well they haven't really. Moving a "Metro" Silverlight project to WinRT requires very little effort at all and can easily be absorbed into an upgrade release schedule - it's not like you have to do a major re-write or change the language or anything.
Oversimplifying once again... If you're speaking of lightweight WP7 apps then you may be right but anything on the desktop isn't going to be like the magical port demos you've seen in the build sessions.
With the W8 lock-in under WinRT if it fails (and specifically the WinRT/Metro side of the house) those that have written apps for it will be in the same spot as the WP7 guys. It won't matter how many copies of W8 are sold (for use of the desktop side or just because that's what ships on new machines) WinRT and it's fledgling apps will fail in their own right if Microsoft can't figure out how to get people demanding WinRT/Metro apps. The Zune, Kin, and now WP7 are great examples of what could happen to WinRT/Metro. It would stand a better chance if they'd back port it to W7 or had forged it as a SL strategy in the first place rather than building this phone OS runtime bolt-on into Windows.
The other piece that's going to be an issue is that the talented developers I've worked with never want to work in "legacy" code. Now that WPF/SL is legacy it's going to be harder to get good devs who want to do the work. Even if WinRT/Metro is a fail there are other places these devs can go now as you've mentioned.