Innovation can only do what is possible. If the font foundries want it to be physically impossible to "steal" the font, there is simply no way to square that with making the font available over the web for the purpose of rendering it, without sacrificing too much usability to make it worth doing. Maybe IE can do it (although even then only dangerously, if you want it to be really secure), but there's no way to make it possible for Firefox, KHTML, and other open source browsers to use the font, but impossible to download it. They're just too open for that. And any solution based on leaving them out in the cold is rapidly becoming infeasible as IE may still have market share, but is very nearly the last closed-source browser of any significant still standing.
Leave the handwaving of "innovative" solutions to impossible problems to the government.
It would probably suffice to develop a standard where only an embedded subset of the font was transmitted by the web server, so you couldn't just download and install the .otf file or equivalent. That would probably save a worthwhile amount of bandwidth with modern fonts and their huge file sizes, too.
As it stands right now, you do, because the font foundries won't stand for anything less. If they were happy with probabilities they would have been in the game a long time ago.
Note this is a social thing, not a technical thing.
I think this whole @font-face (1) thing is exactly what is needed to shake some sense into the foundries. Designers will start using the fonts they love, without needing to create a graphic, then the foundries will start freaking out and (hopefully) innovate new, standards-based methods of using a designer-chosen font without sacrificing the usage rights of said font.
As a slightly different argument, using a font via @font-face is no different than using a graphic. Graphics can have usage permissions associated with them just like a font, so why do we need to invent a new means of distributing licensed fonts?