Update indexing documentation for Exponential Rebates

[alex-pakalniskis]

Removes references to Cobb-Douglas, rebate pools, and legacy allocation states. Tagging @tmigone for visibility.

Thanks 😄

[azf20]

Nice! I think this is good from a "removing prior references" standpoint, I think maybe there could be a short explanation within the docs themselves (vs. always linking to the GIP) to explain why there is a rebate function at all, and in laymans terms how the exponential rebate function works.

[azf20]

I think it might be worth providing a bit more of a legible explanation of what the rebate function is for (essentially ensuring that indexers have sufficient stake allocated when claiming a given amount of query fees)

[howardheaton]

Agreed. To be explicit, after "to an exponential rebate function.", you could add:

This aim of using this rebate function is to ensure indexers achieve the best outcome by faithfully serving queries. This works by incentivizing Indexers to allocate a large amount of stake (which can be slashed for erring when serving a query) relative to the amount of query fees they may collect.

[tmigone]

LGTM, agree with Adam's point on adding a brief explainer on the new mechanism, maybe in the section where we link to the GIP we could add one sentence or two kind of a TL;DR for the GIP. @howardheaton might be able to help?

[howardheaton]

@tmigone, I agree. It's a bit long, but how about this for a tl;dr?

The exponential rebates function is proposed as a way to ensure indexers achieve the best outcome by faithfully serving queries. It works by incentivizing Indexers to allocate a large amount of stake (which can be slashed for erring when serving a query) relative to the amount of query fees they may collect.