You misinterpreted some things
That isn't what he meant. He meant he would reject the mod, not the donors
Clarification never hurts. Essentially then, priority is given to mod authors who opt for a smaller percentage of shares and thereby increasing the potential payout for charity and himself. Correct? Again, not implying anything, just making sure the idea is clear and not misunderstood. A mod author that wants 80% to himself, 15% to charity and 5% to @LivingAngryCheese would get greater priority than an author that wanted 90% for himself, 6% for charity, and 4% for @LivingAngryCheese. (going by the statement about charity getting more, otherwise just pulling numbers out of the air)
He meant the user could turn it off client-side
What do you mean the user could turn it off client-side? Why on earth would a user want to disable their own perks? If you mean a mod-author can toggle perks on and off on a per-user basis whenever they wish, for whatever reason they choose, I still maintain that is a slippery slope.
As a side note, I think I got my dander slightly ruffled with the reference to the community being "snakes". As someone earlier in the forum pointed out when referring to the Ex Nihilo 2 funding effort, one should also not ignore the fact that a large portion of the target audience does not necessarily have disposable income, whether due to young age, limited or no income, or some other reason. The distinction wasn't really made.
I don't think the idea that mod/content creators should be able to request compensation for their mod is a bad one if people want to donate. I personally think a simple framework and associated API could be written that allow just this sort of "added perks" thing for modders to use without all the extra bells and whistles of a website and a middleman. If a mod author uses that API to toggle optional "cosmetic only" content and it doesn't inherently violate the Minecraft EULA, that's way less complicated than what it sounds like is proposed above.
In regards to "donations", like everything else ("that's why we can never have nice things!") some will take it to the extreme, and push the limits until they bleed. The "pay-to-win" types pretty much covered that base. Regardless of your stance on the topic, it is what it is, and that contributes to the general atmosphere whenever the topic is brought up.
I think Patreon provides a pretty decent alternative for several modders, and though I'm sure Patreon gets a chunk of those donations too, it gives users a personalized portal to each mod author, and provides various social gizmos and sparkly-things to incentivize donors to keep donating. This, combined with the framework/API idea I mentioned up above, would seem to satisfy almost everything a mod author would want, short of a potentially larger cut of the donations.
What
@LivingAngryCheese needs to propose is an offering that brings something of significant tangible benefit to both the mod author and the donor. Capes can (and have) been done. Per-mod, per-user perks are either entirely client-side ("Ooh, I look like a sparkle pony!" -- "Not on my screen. You just look like Steve.") or they require other clients/servers to also have the mod installed. The site needs to then communicate with each client/server on some sort of semi-regular basis to communicate which users have what perks. This needs to be fairly frequent and consistent, or the "toggle" mentioned above could have a significant delay before the change goes into effect.
If this requires others to have the mod as well for the "perk" to have any significance, you then need to provide some sort of incentive for people who have no interest in the mod themselves to have it installed anyway so Mr. Sparkle Pony can look that way to people other than just himself. A modder building this into their own mod already provides a reason for their mod to exist on a server or client.
@LivingAngryCheese needs to have one as well. See the difference?
As I said at the beginning, I am not trying to disparage the idea. I'm trying to point out possible deficiencies in the concept. If this results in a better idea, great. If not, no harm done.
Edit: Minor quoting snafu.