Let's be clear on the limits of the study you're quoting -- the
total effect size was a tiny, tiny fraction a percent for the surgical mask arm (.09%), and barely reached a reasonable level of statistical significance (p=0.043). There was no statistically significant effect size for cloth masks.
No matter how you look at this study -- if you actually read it [1], and don't just read press releases -- it was not a strong result. Moreover, after the raw data for the study was released, there were a number of legitimate critiques that call into question the primary result. For example, this one [2]:
> In this short note we re-analyze the data from this trial using standard non-parametric paired statistics tests on treatment/village pairs. With this approach, we find that behavioral outcomes like physical distancing are highly significant, while the primary outcome of the study is not. Importantly, we find that the behavior of unblinded staff when enrolling study participants is one of the most highly significant differences between treatment and control groups, contributing to a significant imbalance in denominators between treatment and control groups.
Even if we accept the claims of the original paper at face value, the fact that you're citing this study as some kind of definitive rebuttal is, deeply ironically, a perfect example of the OP's critique: the highest quality evidence for mask mandates is simply and objectively weak, but most casual observers believe the opposite to be true.
Public health has exaggerated (or outright lied about) evidence quality throughout the Covid-19 pandemic. Is that the reason for this Polio event? Likely not, but if we're casting about for secondary or tertiary reasons, a drop in trust in public health due to their own trivially falsified propaganda wouldn't be the worst place to start.
[1] https://www.poverty-action.org/sites/default/files/publicati...
[2] https://arxiv.org/abs/2112.01296