Some of the things we plan to help them learn: corporate equity, pitching VCs, negotiating investments, dealing with investors, handling PR, user testing, user context research, product management, UI design, A/B testing, running production systems at scale, software processes, and engineering management.
We definitely can't offer them more money than Microsoft or Google can, but we sure can teach them a lot more.
Direct involvement in financing is, as you expected, hard. But we try to be very open. Everybody knows the burn rate and the terms we took money under. Everybody regularly gets updated on dealings with our investors. We also have a regular brown bag series. A couple months back, my co-founder did one on venture financing, covering both the basics and our specific fundraising effort. He did another one recently on the competitive landscape. And he's an awfully good presenter; you can see a 5-minute talk he did recently on our approach to testing product ideas:
So you're right, the coder who would be willing to consider such early involvement in a project is probably fairly entrepreneurial and will lose out if he is taken on as anything other than an equal co-founder.
This has been my feeling previously, but this post and the comments helped make it concrete.
Edit: I realize this is a somewhat heretical idea around here, but the difference in stress when it's your life savings in the mix and your responsibility to make payroll is quite substantial, and being an early employee gives you a window into the process without having to push all your chips to the middle of the table on your first go around.