Small business failure rates: well, for starters, the usually-quoted numbers have been updated as of late. A quick Google shows that most sources are converging on around a 60% success rate for small businesses at the moment. (I think that's actually unrealistically high - I'd estimate more like 40% and guess they aren't accurately polling online businesses, but hey.)
Secondly, there's a significant difference between small business failure rates and small businessperson failure rates. As I mentioned above, most people's first businesses will fail. That will obviously skew the numbers in favour of failure: indeed, given that it's surprising the survival rates are as high as they are.
You're correct that it's (sometimes) easier to build a successful business after you've already built one. However, you don't mention the secondary point, which is that it's also much easier to build a successful business after you've already tried really hard and failed at one.
Most people can not afford a failed business. If you are looking at repeated tries, you are taking a huge amount of survivorship bias already.
For example, infoproducts (or any kind of digital product) are very cheap to start up. They take time, but not necessarily a lot of money.
A successful lifestyle business takes a lot of learned knowledge, and by virtue of building a single one you've had more success than most. Just beating the odds on one was your survival, once you did that you had a lot of advantages.
Bigger serial entrepreneurs often have similar patterns. One success allow you to make another which allows you to make another, all with varying degrees of success and with failure effectively subsidized. See Mark Cuban, Ev Williams, Pewter Thiel.