But, we could be wrong! We'll see.
I suspect that, despite the potential for abuse, you won't actually see very much actual abuse.
You can't obviously measure it on number of lines changed or something because that doesn't reflect the importance of a change, but maybe just say "okay, I can merge this but you get only 0.0001% because it's just a minor typo" or something like that. And if somebody is trying to abuse the system then just give them 0.
A better way would be to set up issues with bounties (what is already mentioned in the article) so the actual developers can decide the worth of contributions before the contributions are actually made, in order not to do favoritism.
Or maybe I didn't properly understand how it works?
One might argue that initially those problems would fix themselves as people would grab for the lowly hanging fruit, and that's fine, but I could see someone refactoring segments of code for no realistic reason, but "selling" it as an improvement, for the explicit goal of collecting the money.
Having a gatekeeper prevents this somewhat, but this company seems more interested in some kind of idealistic anarchy.
[.] Finally! We'll be having people fighting each other for a more up-to-date documentation!
https://news.ycombinator.com/item?id=6882374
(SFErik had nothing to do with the bitcoin stuff; someone else put up the money and started it for his repo.)