readme download counts seem inflated, tracking real usage

[priyam0k]

Our readme download badges seem pretty inflated compared to actual human users.

Most of these spikes are likely coming from automated ci/cd runners, mirrors (bandersnatch), and bots rather than manual actuarial installs.

With upcoming releases (0.11.0 and moving towards 1.0.0), it might be helpful to have a cleaner picture of real community adoption...

[kennethshsu]

Ya... Do you have ideas on where to pull the "more real" numbers?

[priyam0k]

Yes, I have an idea working on this. Posting soon in the parent issue #942 . This keeps coming to mind after this was raised in our meeting