There should be no such thing as a "launch day". You should start with one user before you wrote a line of code, then write a mockup and show it to the user. Keep adding code, features and users as you go.
Watch this for details: http://ecorner.stanford.edu/authorMaterialInfo.html?mid=2056...
Good point. I'd like to add though that while there are merits to launching early and often, you only get one shot at delivering the right first impression. This is especially true if you've lined up the "press" to talk about your app, only for it to receive "bad press" because it's not ready or crash under the heavy load.
I think it was Paul Buchheit who said "launch your product if it's better than anything else out there". To me, it means launch when it's ready but not before.
This is bang on target for me as I've been a real "tightwad" lately. Instead of focusing on the outcomes, I've been counting the pennies. It's tough when you're leaking your own money and that's before you've even launched anything! I think it is just something I need to adjust to - that in order to make money, one must first spend it.
I've also realised that marketing early is so important. I've been getting more active on Twitter lately and though I have a decent blog (been on Digg front page and Lifehacker), I haven't given it the love it deserve all year. I wish I had ramped these two promotion avenues up 6 months ago. This way I'd have a ready audience on launch day.
I thought I'd echo your reply for the other budding entrepreneurs out there. Don't be afraid to spend money! Start creating your story and marketing your brand early!