I still disagree on this (but wouldn't downvote you for expressing that opinion).
I now design and review systems with the assumption that the GPA (global passive adversary) is real. It's not a political thing; it's an observation of technical reality.
To explain why that is a shift in thinking, note that basically every web-site password reset mechanism in the world (apart from those that employ 2FA) is broken in this scenario.
Sensible people cannot expect Tor to provide the fig-leaf of safety it seemed like it offered.
GPA was not a default assumption in threat models before.