Grants/Sponsorship still exist.

Andrew is not arguing against people paying for specific features/work. He is arguing against 1) not coordinating that work with the core team 2) competitive 'first correct submission wins'

From by my own experience, you'll have to review/test the boring task from a one-time contributor, which is also not fun. You don't need to do that as often if the contributor is going to stay, but you're not attracting this sort of contributions with bounties.

The last point is also similar. If I get a significant contribution in area where I'm not competent, I might take it, but is it going to maintain or fix problems to it in the long term? I'd rather not include a feature which I know I can't maintain long term.

In my experience its rare for there actually to be boring useful tasks in open source. Everything is interesting to someone.

Sometimes things are boring because nobody really is interested in the results, but i think that is a diferent kind of boring. Sometimes maintainers are overworked and cant do everything, but that is also a different sort of problem.