I appreciate the concerns, but don't like the proposed implementation at all. Most obviously, nothing prevents bad actors from creating multiple accounts. They can use burner e-maile and phone numbers, spoof IP addresses, and so on to circumvent most basic prevention methods, too. Reputation paints could be easily manipulated.

Free starter HP would drain faster than it does now, talking the economy further. I would enforce a DHP HP delegation program for new accounts, but again, people could easily create fake accounts and promptly farm votes for spam and plagiarism.

I think the KE metric is part of a real solution, but like high reputation, a low KE ratio doesn't inherently signal virtue.

I think the real issue is trying to place a numerical value on reputation in the first place. Perhaps it should be removed entirely, or its effect on votes and how it influences reputation should be changed. As it stands now, low reputation accounts cannot affect high reputation at all.

Maybe a weighted system is needed instead, combining age, KE, reputation, upvote history, and other factors in a complex Genuine Engagement Metric formula would be at least more difficult to game by bad actors? If someone has a lower rep, but also a lower KE and higher GEM, they could affect higher rep accounts? I really don't know what the best solution would be, though.