The idea of war tribute / reparations for peace sounds good on paper, but I really don't see how it helps address the simulation's current issues. If anything, the winner will be the recipient of (perhaps juicy) daily payments, with which to further upgrade their army and stomp someone else - until the payments end. From the perspective of simulation health, this seems to me a strictly inferior approach to the simple solution from 1.3: forcing a 25-day truce without any monetary exchange, so the loser can lick their wounds and maybe come back stronger next time.
The problem is game-ending snowballs. Winning as a snowball is dull. Even losing to a snowball is dull: sure, you might turn the tide, but they've probably crushed most other factions already into dust, so your playthrough's scope for an exciting emergent narrative is pretty much over. So, you restart.
What should be some of the priorities for keeping the simulation engaging?
- Internal factors within a faction, to dent snowballing: think "Corruption" mechanic from something like Total War: Three Kingdoms, or indeed earlier iterations of Civ (e.g. Civ3). In its simplest form, the more fiefs a kingdom owns, the greater the percentage of tax that goes to waste in corruption or logistical overhead. Assuming every AI faction obeys proper economic rules (I am led to believe they do, from developer posts in these forums), that corruption percentage (adequately tuned) should significantly dent the power creep of larger factions. I never much liked Corruption in TW:3K, but judged purely on the gameplay impact, it did exactly what it needed to do: it put the brakes on a faction's ability to spiral upwards. Bannerlord needs band aids to fix simulation health, this is one of the simplest ones I can think of.
I can imagine other internal factors worth pursuing: claimants to the throne like in Warband, or internal revolts spawning new factions when one grows too large. These are more complex to tune correctly, and done wrong will also lead to farcical simulations. Definitely would love to see these features eventually, but development time dictates that this should be looked at later.
- Diplomatic factors between factions, to dent snowballing: basic diplomacy AI capable of recognising rising threats and acting out of self-preservation. Think "Great Power Aversion" from Total War games or "Coalition" mechanic from EU4: a system so factions simultaneously recognise the need to contain an out-of-control neighbour, by ceasing their own hostilities indefinitely and banding together to ensure containment. In particular, EU4 modelled "Aggressive Expansion" as a metric of a faction's belligerence - a metric that degraded slowly, over many years. A faction might lock in territorial gains over time, but it would be years before their past aggression was water under the bridge. If they expanded too fast, took too many fiefs too quickly, the coalition mechanic would kick in and cut them down to size. The political borders of Calradia would change slowly over years, even decades. This 'slow drift' has to be baked into the simulation, otherwise the entire legacy/inheritance system is pointless and misleading.
- "Events" or externalities to dent snowballing: this could be any of a list of events designed to draw the attention and resources of the snowballer. Perhaps their kingdom has become too affluent: bandit lords from all over Calradia converge on their kingdom, forming great bandit armies (with maybe unique T6 loot, as a further incentive to explore this lategame experience) to ravage their villages and siege their fiefs. Perhaps a great Crusade is called, by a faction opposed to the snowballer: well-supplied, high-tier armies of the snowballer's nemesis converge on their lands.
A flipside of these anti-snowball measures: the game should no longer punish small kingdoms. In grand strategy, you do not mind the small neighbours! They are an irrelevance, but may make themselves useful if they align against your enemies. Likewise here, a player-made, small kingdom should not be the immediate target of all war declarations. Their "Great Power Aversion" is minimal, their "Aggressive Expansion", initially at least, is zero. If the player has carefully lain the groundwork (built up positive relations with neighbours + high enough Clan Rank that kingship does not ruffle feathers), their early kingdom life should be a breeze, with widespread hostility only manifesting once their expansionism has soured those relations.