The biggest problem I have with this thinking is that it's a false dichotomy - it's not "Salary or startup" at all, and the fact that most young software professionals see those 2 as their only options saddens me.

There are plenty of other routes to go - freelancing, for one, which done well can give both higher earning potential, as well as more flexibility in terms of how much money you can earn. An effective altruist may well decide to work slightly longer hours for more money, something that isn't as possible in a normal salaried position.

Another option is starting a consulting practice (e.g. what I do), again, I'm still early in this game but it may well offer a higher expected value. (Or not - I don't have figures).

Another option is to start a small bootstrapped product company. If your aim is simply to maximise your personal cash in order to donate it, small non-VC product companies have a lot of merit - I think (again, no real proof and totally anecdotally) that their expected value is higher. And if you're into maximising expected value irregardless of how, this may make sense, even if a small product may have less chance of "changing the world" (one of the reasons that it's more likely to succeed).

I'm just saying, there are plenty of options besides being a salaried employee or taking VC money, and money of them seem to be better, on the whole, especially for young software professionals who could use to run a real business and gain valuable information on how running a business looks like.