All of my previous concerns have been addressed that can be. Thank you. Some of the confounds cannot be correct post hoc such as between present and future uncertainty and magnitude. Likewise, since pulling number from models in real-time (give or take), the magnitudes of uncertainty are not constants.

I advocated for the inclusion of the cross-cultural, and I agree that random-effects are a good idea. I would start with the simplest model possible. More complex models, like with random slopes, are not necessary to test if country-wise differences matter. There seemed to be some trouble fitting the more complex models confirming that the simpler (ie random intercept) models are a safer choice.
What could help provide a general result for the countrywise analysis is something like a rank-order correlation of the three outcomes by mean country score. How strong is the negative correlation between perceived uncertainty and trustworthiness?
Do you believe the output from the countries models are robust effects? Is it reasonable to expect similar countries to have similar outcomes. Culturally similar meaning (UK & France, UK & US, Italy and Spain, etc). Why are China and Japan so different on these dimensions? Why are Sweden, Australia, N Korea, etc so central on these outcomes?
Some minor comments about the new Figure 1.
What are the effect sizes of marginal means (maybe a standardized mean difference) for panels A – C of Figure 1? Do these effects have practical significance?
Shouldn’t ‘Perceived Uncertainty’ be reverse coded to compare with ‘Trustworthiness?’
For panels G & H is there any way to get more of a vertical space between countries to help divide them more clearly?