Oh... there's *plenty* of room in Sprint's 800Mhz spectrum for W-CDMA... or at least, there WILL be, after T-Mobile makes Sprint gamble it all on the merger, the feds shoot it down, and T-Mobile walks away with it. :-D
Don't feel too bad for Sprint, though... whatever they lose, Softbank will just turn around and replace with new 600Mhz spectrum anyway.
Food for thought: the only country on earth whose mobile phone market is more restricted, locked down, gimped, proprietary, and basically run in ways that make AT&T and Verizon look like poster children for open hardware and Cyanogen is... Japan. Nintendo -- the most dominant videogame company on earth just a couple of years ago -- was blinded by its messed-up home market and concluded that phones were a hopeless lost cause. Sony was, too... until it bought Ericsson. Softbank wants to replicate its "success" in the US.
Imagine, for a moment, if Sprint were in charge of T-Mobile and got to dictate policy going forward during a transition... and did something completely evil, like disabling new, non-grandfathered T-Mobile SIM cards from working in non-T-Mobile-branded phones (whose ESN wasn't in Sprint's database).
A Sprint-Tmobile merger would be a tragedy because it would eliminate the only lifeboat to semi-freedom we *have* in the United States. And that's precisely why the feds will find a way to shoot down the merger. T-Mobile has done a great job of casting itself as the white knight who's here to save Americans from being exploited by Sprint, AT&T, and Verizon. The feds won't care whether or not a larger Sprint would be "more competitive" with AT&T and Verizon... it's going to care whether a Sprint-owned T-Mobile will still be "kind of unofficially non-blatantly-evil".
There's another reason why I believe the merger will never happen, even if the feds won't stop it: Google. T-Mobile has been Google's favorite bully pulpit since day one. It's been the one American network where Google has never had to kneel before anyone and play "Mother, May I?" I firmly believe that if push came to shove, Google will show up at the last minute and buy T-Mobile just to keep it from being swallowed by Sprint, Verizon, and AT&T. It'll buy it just to make sure it never has to beg for permission to do anything.
As far as 1xRTT is concerned, 1xRTT will be the LAST legacy service Sprint shuts down. 1xRTT will be available for YEARS after the last site with EVDO repurposes its spectrum. When it comes to wide open rural areas that occasionally have an influx of tourists, 1xRTT has some real advantages over W-CDMA (mainly, it can accommodate more simultaneous users in narrower channels, instead of just a few users splattering high-powered noise across huge chunks of the band). Two dozen W-CDMA users with directional antennas & high-power picocells can't hear each other's uplinks, but the tower can hear them all stomping on each other. 1xRTT is a convenient way to separate them by space into smaller chunks of spectrum than W-CDMA would allow.