That is the tricky thing about basing predictions on historic trends. IIRC, is was Arab spring in some countries that broke the rule (and that "rule" was calculated based on data from 1900-2006). I'm also curious how Nepal and Madagascar stack up compared to historic trends.
Other aspects of Chenoweths research (such as the importance of "converting" certain factions such as police or military - or at least getting them to be neutral) are important as well but don't get as much traction as the clickbaity 3.5% number does
ETA: Nepal's discord had about 100k members, and it's population is 29 million. So that was well below 3.5%. Like No Kings, Nepals protests were decentralized, which is potentially a factor compared to historic protests which tend to be focused on a single location or march
The tricky part about a general strike is that you dont have much room for do-overs if it turns out you don't have the numbers you need. Protests are largely symbolic, but are also a good way for organizers to gauge how willing/able folks are to engage in more drastic action.
If a general strike is what you think the movement should go towards, I recommend your next steps be strengthening community support systems - food banks, free clinics, shelters, etc - since these things will be needed to support a general strike (folks will choose feeding their children over political action, so make sure they can accomplish both). In the mean time, it also helps out folks who have been hit the hardest by republican and trump cuts to services.
I highly recommend folks read up on some of ICNC's stuff, since they focus on the practical side of nonviolent conflict.