IE7 may replace firefox for some users, but not those who care about proper rendering of web pages. Despite years of promises, IE7 still does not meet CSS rendering standards, among other rendering problems. There is also the matter of security. I'll be the first to admit that M$ has made great strides in security over the last few years (compared to their previous security priorities), but many of the advances they've made are only halfway to where they need to go. IE7 will by default disable many ActiveX components, but not all, and the average user doesn't understand enough about the horrendous security vulnerabilities of ActiveX to stop them from re enabling them...the first time the user sees the popup that advises them ActiveX is required to be enabled, they'll click yes...opening the browser right back up again. Further to that (and noted in the article), only in Vista (delayed yet again) will the browser be isolated from the OS. Most IE7 users will run XP...as with any new OS, the intial uptake will be slow, especially given the recommended hardware requirements for Vista. On XP, IE7 is still integrated...a hole in the browser is a tunnel to the OS.
Finally, IE7 will still be proprietary software...for power browser users, this is not good enough...there are simply not enough plugins for IE for power users, as M$ doesn't want anyone but them building them. For power browser users, the lack of plugins will ensure they still use ff (if they do now).
Overall, I expect to certainly see some FF users switch back to IE, but not enought to kill FF...it will be those interested in eye candy as opposed to true functionality.
Having said all that, I still wish FF would concentrate more on plugging memory leaks than new themes!