These numbers are interesting, but they don't exactly reflect either the Colymar militia or the Jonstown tribes, which may be indicative of both groups being slight outliers.
The average clan post would posit that around a third of a clan's adult population are able to fight, with 90% of those being militia rather than dedicated warriors. So about 30% of a clan is militia, and lets assume that these numbers scale up reasonably well to the tribal level. The GM Screen Pack actually differentiates between dedicated warriors and militia for its army. 2200 militia and 300 warriors. It also helpfully tells us that its adult population is 5300. So that's about 47% of the tribe being able to be called upon to fight, with 41% of the population being militia, which is a significant margin more than the clan average. It also has a slightly higher percentage of full time warriors making up its army than the average; 12% compared to the average of 10%. So the Colymar does look to fight more often.
We also shouldn't ignore that the Colymar has a massive difference in its adult:children ratio compared to the average clan numbers. Two thirds of a clan are adults, whereas the Colymar adult population is only about 43% of the population. The book does draw attention to the large number of children and young adults, also mentioning population damage done by the Great Winter that has since recovered. Considering the Colymar can't be supplied by a city like the other tribes, the Great Winter probably hit them harder, but I could also imagine it may have inspired them to resort to banditry or leaving to rebel against the Lunars, which may explain why they have more warriors than the average,
Now the Starter Set is less explicit about what's going on with the Jonstown tribes' population of adults, but we can suss out that the total population of the tribes is 26,200 I'll ignore Jonstown itself for this analysis since it will be a bit different and we can separate its militia handily as the direct Jonstown contribution is given). Assuming the Jonstown tribes follow the clan average, that's about 17,500 adults, which going by the average gives a number around 5800 for the army. Now that is vastly smaller than the 2500 that is being stated in the Starter Set. The book does admittedly refer to them explicitly as militia so there might be some full time warriors being excluded from these numbers, although I'm inclined to think they are included but are just not named explicitly for the sake of simplicity since those numbers are definitely important to understanding the area. But that's a lot of numbers to make up when using the average clan numbers, and using the average clan post to say that the 5800 includes the full time warriors would still make the militia total at 5220 or so. So the conclusion I'm drawing here is that either that the average clan numbers can't be used at a tribal level (which I feel they probably can to some degree) or that the Jonstown tribes are very atypical tribes, like the Colymar are but in the other direction, either having a much smaller fighting force than normal or having an as of now unmentioned in RQG force that exists outside of the militia.