A possible explanation of your data "loss" is the merging of the separate CIT and CGT character tables. It's possible one was out of sync with the other, and the incorrect one won out. I didn't do that merge so I can't say what exact logic was used, but that seems the likely cause.
Another important note is that anything that was being "tracked" during the downtime didn't make it to the new database, that's a known issue.
Search is coming back, we're still working on the details of that. We're trying to weigh functionality, usefulness, and privacy differently than before, but it's definitely high on the list.
Yes, we will be making public APIs for you all to get data back out for your own sites and projects. Whether that ends up being .NET webservices or straight XML feeds, that's yet to be determined.