Do you actually struggle with it though? I have to ask because you seem to be able to believe such high numbers without any skepticism, I don't think one can entertain ancient historians as a whole when they come up with such numbers.
It's a multifaceted question, and I think you're doing a grave injustice to the topic by reducing it to believing ancient historians or not. A person can accept evidence that supports large ancient armies while still not understanding why later armies weren't as large.
But if it's all contingent on choice why would anyone stop doing it? Why would someone purposefully limit its manpower and forces? Ultimately there must be a financial, logisticial and strategical reason to do so, it can't be all political and economic structures, otherwise I'd argue we would see those structures converge towards the most military capable states if structures alone allow for such huge armies to be raised. Anything else, like individual unit quality, would pale in confront to the ability of one state to mobilize so many more men. Also I think we shouldn't seek easy explanations such as the elimination of the full freemen levy, because we would actually need to see if such a thing is actually possible to begin with in settled societies and whether it was ever actually used, especially in societies with large amount of freemen.
What's your standard of proof for establishing that universal levies existed in antiquity? Right now, it really seems like you're flat out unwilling to consider any rates of mobilization over 2-3%, no matter how many historical sources widely accepted (with caveats) by qualified modern analysts one produces. Regarding the elimination of universal service, sometimes the good of the elites isn't the good of the community as a whole; if elites have to concede power to common people to improve their collective warmaking ability, it's a real question whether they'll go through with it. Maybe the common people don't want to be compelled to pay taxes and fight for the state. There's a lot of potential explanations that don't rely on simple voluntary choice; that's what we should be talking about, not whether to throw out every ancient historian's figures.
The Romans spent plenty on public works, specifically in the imperial era they also had subsidies of various kinds, internal and external. Even the professional Roman army of the late principate didn't compromise all of the state expenditure(it amounted to about 60-65%) and left a lot of room for other stuff.
https://www.princeton.edu/~pswpc/pdfs/scheidel/041201.pdf
In the Mid Republic, the budget spent on war across 43 years averaged a bit under 80% (see Michael James Taylor's dissertation
Manpower, Finance, and the Rise of Rome), which included some years of much lower military activity than others; in years of peak activity, it's only natural for this share to encompass almost all the annual revenue available.
And modern armies have a lot better technology, demographics and bureaucratic structures, if you want to compared relatively untrained levies then the Romans still levied armies on the relative scale of Napoleonic mass levy(well quite bigger apparently, 2-3% vs 5% or 10-14%).
The extreme examples from the modern period are fairly comparable. Sweden during the Great Northern War and Prussia during the Seven Years and Napoleonic Wars were able to raise 6-7% of their population; the mostly non-industrial Confederacy in the ACW had 11-12% mobilization, despite the much more sophisticated military machine at work.
Which census exactly? Also I'm not sure why we have to entertain the idea that Hannibal had an actual 1:30 ratio disadvantage and that the Italians could mobilize upwards of the 10% of their population, at some point we must be able to say that X historian is wrong regardless of what supposedly good sources he had.
The censuses of the Mid Republic through the early Principate we have through Livy; with a couple exceptions (source critics suspect corrupt manuscripts), they record 200-270,000 citizens in the late 3rd early 2nd century. Also, I think you misread that passage from Polybios; we wasn't fighting 770,000 men at once, that was the pool of
assidui the Romans could draw on, while if you add up the forces they had under arms for active campaigning, the total is a bit over 250,000. At the beginning of the war, the Romans raised about 60,000 men; one army of 20,000 fought in Spain, while the other, recalled from Sicily, joined the local forces for a total of 40,000; at the same time, Hannibal had recruited the local Gauls to reinforce his army to about 40,000. Depending on what estimates you use, the Romans raised about 175,000 men at their peak, as did Carthage; most of the Romans were in Italy containing Hannibal, while Hannibal's brothers in Spain were defending their manpower base with the bulk of Carthage's forces. None of the comparative balance or distribution of forces breaks probability; each side had a relatively defensive posture, protecting their manpower bases with forces close to hand, while attacking the enemy's base of power. If the Romans began the war less mobilized than they could have been, that's hardly unprecedented; it's very easy to misjudge the strength of an enemy, and false concepts of economy have proven difficult to get rid of.
Archaeological evidence such as? I'm being repetitive but you should link when you mention such things. Also what does it mean, "able to bear"? I don't doubt there were 50k male adults in the regions.
Douwe Yntema's analysis ("Polybius and the Field Survey Evidence from Apulia" in
People, Land, and Politics: Demographic Developments and the Transformation of Roman Italy 300 BC–AD 14) of things like settlement size, density, and distribution accords with Polybios's population data, which was derived from the same source as the army numbers; it's hardly credible that the same official records would have an accurate survey of population, but wildly inaccurate numbers for their own armies, which were after all their priority.
Depends on the nature of war but it shouldn't go much above 2-3% for non-local conflicts for agricultural based countries. That's just a general figure, obviously specific cases the ceiling could be higher or lower(though not to the levels ancient historian go)
And how do you divine this figure? Actually, don't tell me, because if you can turn this into a book and apply this to the study of ancient warfare, disproving a century of analysis by the best and brightest the academy has to offer, you're on the track to becoming one of the greatest ever historians of ancient warfare.