First of all, get a new server. It takes only a few minutes with most hosting providers. DNS updates will be a while to re-map but maybe your "old" provider can help you with a temporary redirection. Get this up and running as fast a possible, even if you use a temporary domain name. Go, do it. Now.
Then, the post game analysis. Server apocalypses happen. There is a deeper lesson here.
On the technical side, there are some things you can do to prevent this kind of failure. Running fsck for hours on a live server is really inexcusable both from a technical as well as a service point of view. Most small hosting providers simply rent some rack space, put in their standard Dell server unit and call it a day. This way lies madness, total data loss and possibly the early seeds of an anxiety disorder. I'd suggest you look into inherently redundant hosting offers (like MediaTemple's grid service) or some high-availability virtual hosting (like Amazon EC2).
But the core of the issue is unrelated to this technical stuff. Telling everyone at once was a mistake. There is a reason why most projects start out gradually, starting with a beta phase when everyone still has a lot of compassion for bugs and crashes. It's tempting to send invites to all possible contacts once you're finished, but it really has no upside because you're introducing a single point of failure. I'm not even talking about having only one server and no backup. In this case, the blitzkrieg launch date itself was that single point of failure. Even if the server hadn't crashed and burned, your app is very likely to still have bugs and kinks to work out. Bringing potential clients in one by one would have given you the opportunity to collect valuable feedback so you can show a more refined version to people who come in later.
When it comes to handling this issue from the CR side: don't send out apologies for the outage. Instead, check back with high-potential contacts in a few days and ask "did you have a chance to look at my new project?" and when they reply they couldn't access the server, invite them to try again and concede there were some outages but the problem is fixed now. My guess is, you'll be surprised how few people were actually affected and noticed that the server was gone.
In my opinion, it's unlikely that your bad luck actually ruined anything. Most people don't look at random invites in their mail right away. If they're really interested in the product and/or know you personally, they'll check back again later even if your site is offline at first. Interested people will still look at your stuff. Chances are, those candidates for your first mass invite will not even turn out to be your most valuable clients. Nothing is lost.
