The Magic Answer Box people are demanding that they be trusted with nuance and detail and yet continuously demonstrate Magic Answer Box thinking in the face of nuanced and detailed information.
You did get nuanced messaging
"Yeah well... Mandela effect!"
We've also had the CDC telling us the vaccines are safe and effective, when neither are strictly true. They may be less dangerous than getting a severe case of COVID, but they don't confer immunity, and they don't stop you from spreading it, in direct contradiction to other statements they've made.
Is it any wonder that people misremember who exactly told them that masks did or don't work, and when they heard it?
"Vaccine is effective" does not and has never meant "confers 100% protection from infection."
The vaccines remain both safe and effective in the same way that e.g. Tylenol is safe and effective. If you want the fully nuanced definition of these statements mean in the Tylenol case or the vaccine case, you can literally look up the data from the clinical trials and see it. Of course your rebuttal to this will be, "well, I can find X dataset with real world evidence of Y claim" and I'll say: well, there's a reason real world evidence is (counterintuitively, given the name) not considered to be very good evidence in pharma world. That's because it's incredibly hard to isolate variables in analysis of real world data, even assuming the data exists in an analyzable format (which it generally doesn't - also contrary to most people's expectations).
Hmm no, it's no wonder that people misremember who exactly told them what. We have a shitshow of a media ecosystem and you have "contrarians" like yourself who are pumping nonstop garbage like "CDC said X" into it when CDC did not, in fact, say X.
Mistakes were certainly made and everyone was learning as we went, but your line of inquiry actually points pretty solidly away from the solutions we actually need: 1) there should NEVER be PPE shortages in the richest country on earth, 2) there should RARELY be testing capacity shortages and sitting presidents should NEVER deliberately slow down pathogen testing, and 3) there should be much better data infrastructure that allows us to understand all of these things in real time.
Some blame for each of these lays at CDC feet, some at FDA, some at NIH, and so on, but fixing these requires an entirely different conversation than the one we're having.
We don't know this about condoms, either.
What we probably know is that if 100% of the population wore masks 100% correctly 100% of the time and kept 100% of their prior behavior, it would probably reduce transmission more than 0% for some pathogens.
This is another way of saying, “we don’t really know,” and it gives you an idea of why it’s hard to figure it out.
The ambush on “masks don’t work” was because it seems almost certainly the case that masks don’t hurt, and it’s likely the case that if masks have any benefit, that benefit will be proportional to levels of compliance, so if we get to choose between levels of compliance, higher is probably better.
My personal view is that after half the population came to view “no masks” as a part of their political identity, the overall pushiness on masks probably netted out to negative, but this is not blame I can assign to CDC et al.