Historically, this has not been true. Depending on your definition of "rich", very few wealthy people have gotten there by investing in other people's companies where accreditation was required. Most wealth is built by via ownership of your own company, real estate (which is a type of company), or by investing in public markets. These pathways are all still available to you, and almost certainly still represent a better long-term risk-adjusted return than ICOs.

If Oculus had done an ICO instead of a kickstarter, my $400 wouldn't have been merely a preorder. The $400 would've become quite a handy sum by the time Oculus got acquired.

This actually is an interesting point, but it does require some serious cherry-picking. I'd be interested to see the data behind a scenario where all the funded startups for the last decade had been ICOs instead of venture-backed. I doubt it'd look so rosy in that scenario, especially when you consider that many more failures would likely have been funded, based on how easy it is to raise money from an ICO based on nothing other than a pretty website template.