Heh, well I don't know what your experiences have been....But in my experience if one knows what they are doing they can get this done properly and quickly.
Last year I architected and executed a move of a LAMP cluster of roughly 80k domains, for 4.5TB of data, with over 30Mbit of sustained customer traffic with an average of 15GB/hour data delta on the storage volumes(non-log). The move was from Los Angeles to Chicago and took less than 8 hours of downtime. (Be sure it took longer to prepare, but there was no customer impact)
I think one should be able to move a single website, with a single DB instance (or cluster), in a day... :-p Unless of course the architecture has inherent design flaws making it inflexible.
That being said, Calguns.net is free for us.... Until you pay the bills, don't bitch. :-)
I've said it before, and Ill offer again.... if Geovario or Kest wants another perspective on the current problem set I would be more than happy to help where I can...
I work in IT and can tell you it takes longer than you think to move a DB and make wholesale equipment changes.
Additionally making the move in stages so the site is up as long as possible between down times complicates a bit more.
I think they're doing a good job balancing down time and up time. This ain't like copying files on your home PC.
give them a chance to finish the job.
Oh come on.
You know it should never take more than a couple hours (scheduled overnight) to move a large database from one server on to another.
The outages are just WAYYYYYY too common.