Companies and individuals who benefitted from public-good should pledge to tithe 10% of their profit to research - especially that which falls outside of traditional institutions. It is nebulous a priori which research should produce public goods vs private goods, especially if it isn’t focused on creating either patents or papers. Pay-it-forward tithing could enable people to embark on research projects focused purely on reaching whatever the goal of the research is without worrying about either needing to create specific ROI or making sure to adhere to the optics demanded by open source projects or charities.
On top of enabling projects that don’t fit well into traditional boxes of non-profit charity or for-profit investment, a culture of pay-it-forward tithing could decorrelate research funding. The vast majority of research funding comes in the form of grants from either the government or large foundations. Putting aside the question of their ability to allocate money, money coming from a small number of similar organizations with a similar set of constraints on that money (§Philanthropy and Non-Profit Constraints, §Academia Constraints) will lead to more correlated lines of inquiry. Mechanisms to decorrelate research are good because More Uncorrelated Mistakes lead to more success.^1
^1: More, different, funding sources aren’t a magic decorrelation bullet, of course. There will still be trends and hype cycles.