@chess240What you are describing is not a rating system, and I'm quite confident that you are misinterpreting what Mark is trying to say.
Your first quote has absolutely nothing to do with rewards, incentives, psychology - it relates none to the topic at hand. He is talking about how the rating system measures ability, and how it makes assumptions regarding how ability/performance is measured and it's contributing factors; how our models could be improved.
Yes, the rating system is free to be tweaked and improved to more accurately seed players in terms of relative skill. But punishments/incentives are no concern for what is a statistical system intended to be used for measurement (and nothing else). It is not there to pat you on the back and say "well done, you won!"
And Glickman's quote on page 36 is no apology for the rating system, but for the fault of the chess player, "I've hit 2100, I've achieved what I want to achieve. I see no reason to potentially lose this rating so I'll hang my coat and never play again"
Yes, this is a problem which I and many others are guilty of. But Glicko does actually try and compensate for this by increasing the RD of a rating relative to inactivity. (the method by which RD is increased is one of the things Mark is alluding to in your quote)
What you seem to want is gamification, "+50xp Achievement Unlocked!"
A constant reward system simply for being active and involved. But this sort of a system is useless in measuring relative skill.
Achievements have been something we've wanted to add for a long time and I hope to see one day (I think we've been discussing it for at least a couple of years now), but this is no replacement for Glicko.