I honestly can't believe you people are about to throw away one of the biggest conquests we have ever managed to pull off for pocket change when we are getting a bonus to the negotiation roll because ...
(emphasis added to your emphasis)
I can't believe you think we've pulled it off until the ink is dry. Alalkan said no. We know he is willing to fight and die for the sake of his honor. It's so integral to his characterization that I think it is willful blindness to assume one roll can erase it.
I do not believe that this supposed bonus to the roll matters.
1) The negotiator came to the table. He told us what he was not willing to give, and then listed a whole range of concessions that he and his uncle, after watching Methiantese rulers for years (there's your precedents! they know all about our strongarm tactics) could yield. The bonus does not need to make vassalage possible. Rather, I believe the outcome table looks more like a pass/fail check. Any difference in positive outcomes is at most quibbling about the numbers and not the grand scheme. Meanwhile, that leaves 2/6 chance of, yes, getting worse. Those odds are enough that I can sincerely doubt the benefits of massaging good outcomes would outweigh the risks of losing.
2) The bonus is for "Good Terms" treating Parsia
well in spite of our military control, and we first saw it not when we opened our arguments but when we were given the negotiations for a compromise. You seem to think that continuing to ram our plan will always give that bonus. How long before Alalkan decides we're mad and impossible to reason with?
It is the #1 thing we can do to reduce the chance of rebellion later.
I've been thinking about this, and I call bogus.
Case A) The NPCs behave rationally, or at least motivationally according to their own simulated intents. Then the best way to prevent future conflict is to befriend them rather than to bully them.
Case B) The NPCs behave randomly, or at least according to narrative demand. (Haspen has many, many times referred to "rolling" for certain events to happen.) Then it is useless to bully them into [insert safeguards here] because they won't care when they are chosen to be agents of Fate.
I raised a parallel question over Tywynn and Leath. Everybody who wanted Tywynn dead was adamant that being a twin somehow increases the chance of rebellion, but the nearest thing to proof was something like "Well, it's more likely
if Haspen thinks it's more likely." No one ever pointed me to an example(s) that Haspen in fact thinks like that.
In the realm of armies and nations, I have raised the example of Eval as evidence that Haspen rewards our kindnesses ("Good Terms", eh?) with simulated loyalty. Also Dogethe, whom we first met when Yoe I invaded her country. We negotiated faithfully and became friends; even her eventual betrayal failing to stand at length against a
dwarven army was quickly followed by an apology and compensation, which we accepted. Said loyalty may not last forever, about a generation for non-empire NPCs. But imposing our will by force only seems to encourage resistance by force for the same generation.