. . that the functional equivalents are in place for major horizontal categories and that they are good enough.

Linux won't be able to satisfy the PhotoShop or AutoCAD users for a long time. The first is a career application and pratictioners aren't going to be willing to give up any of their hard won experience (and some simply can't give up CMYK). The other has the same experience problem plus the need for 100% file compatibility for extremely complex documents. Fortunately, while these are very important high profile applications, they are niche applications.

What Linux needs is specialty apps that get it in the door, and from which the general apps can be leveraged. I'm sure this is happening, I just don't know where yet.

The biggest boost for Linux on the desktop will be when things have changed enought that Office 97 and Office 2000 are no longer viable and users are offered the choice of moving to registration enforced versions or moving to StarOffice / OpenOffice. A typical 5-user business has 0 to 1 copies of Microsoft Office because it's so expensive.

Once forced to StarOffice / OpenOffice, the big step will have been taken, and the step to Linux just won't look that steep any more. This presumes, of course, that SO/OO will be able to offer the new features that can't be had in Office 97 / 2000. I'm sure Microsoft will do everything they can to see that doesn't happen.