@Sturmflut
Unfun things need the most investment.
The incentive structure around open source encourages the novel (where folks explore new ideas by building them), and the one-off (where folks make things work for themselves).
OSS lacks incentives for the long-tail of feature requests. Examples include Bulletproofing fragile systems to work in more environments (Audio management in Gnome Calls); Necessary security work (e.g., PKCS signing in Geary); or reverse engineering dumb corporate enshittified systems (Exchange).
Basically, going from "Works for me" to "Works for 99% of users" is a huge slog that takes far more time than getting the feature working in the first place. Making matters worse, it's harder to get that big rush of endorphins from fixing the fiftieth minor assumption made about date formats, as compared to the first time the system works.
Effective OSS investment, IMO, needs to focus on paying devs to work on foundational (boring, not-sexy) maintenance and polish work.