Dishonest Tor relay math question - tor-talk is to lazy

[Undescribed Horrific Abuse, One Victim & Survivor of Many]

Back in the days people were used to using mixmaster etc, whereas
others migrated to then-modern p2p systems like i2p [and the word on
the digital street was that "real" anonymous people had custom built
onions that propagated through compromised systems -- and one saw a
lot of those, and there were simple tools all around]

If you look back at the history of mixmaster, it really stagnated at a
point, and it looks now like a struggle of very few devs to keep a
maintained codebase.

I don't know, of course, I was never involved with mail relays, but
it's an interesting parallel to the Tor situation.

Similarly, it is quite hard, nowadays, for me, to find modern
anonymity research papers. As Tor was getting more criticized, a lot
of new research was cropping up.

It used to be that most of the internet was anonymous, really. A lot
of the norms or topics of discussion still assume this, when it has
possibly been no longer true in general.

A bit of a chicken-and-egg situation: you may need to be anonymous to
produce a popular anonymity product that works. Still, they have been
happening.

Any project or paper people would point out nowadays?

Tor still seems quite useful.