I know I can seem very thick, but I still find weird that the "different but intelligible" Romance speeches are actually intelligible
That's historically what happened, at least up to the Xth century when Italian became a really distinct language from the overall Romance speeches that existed before (admittedly old French was quite distinct from the late VIIIth century) : Strasbourg Oaths are a good exemple of this, as it's not really easy to say if it's old French, old Occitan, Gallo-Romance speech, or a more or less artificial combinasion of all of these.
By the classical Middle-Ages, the linguistical divergence continues, but you have still a lot of common features, enough for some exemples of multual intelligibility.
Or, someone that know how to read Occitan generally have good chances to read (again, not that fluently of course) a text in Catalan (altough it's another story for spoken discussion).
On this regard, it wasn't really different from what existed for a long while in the Arabo-Islamic world where, while clearly distinct, two Arabic speakers one from Morrocoo, one from Iraq, could converse (if not as fluently than with their own neighbours)
like I said before to me it sound counterintuitive that even with substrates and superstrates the language will diverge further, I mean before romanization the previous language was a completely different one, if during centuries the area was assimilated to an accent-rich and semi-creolized language why would the language diverge even more?
Again : distance (a bit like distance allowed French Canadian and French metropolitain to diverge in spite of French Canadian being largely representative of some archaic features in French), political distinctivness, popular evolution (without mass schooling and mass media, popular dialects are basically living their own regional life), etc.
I'm under the impression that you treat Roman Empire like it was a current nation-state, that centralized cultural and liguistical features. It didn't. It couldn't.
Look at any other european state before the XIXth : it's a mess of intelligible, related but diverging dialects (at the point you show someone an exemple of Picard dialect compared to French, he's not likely to really understand it). I agree that being used to a dialectal complexity can lead to a better overall understanding, but as long you have decentralized cultural core including in institutions (schools, chanceries, etc.) you won't have much odds to prevent popular speeches to live their own.
Again, that's what happened historically, with relatively close Romance speeches that diverged further and further for centuries at the point forming different languages. I mean,it's not being counter-intuitive or not, that's a fact. That was happening.
But Chinese languages even with political unity are super divergent, I personally think ones needs some kind of modern transportation or mass communication system to have such a large country speak a lingua france understandable from each extremity of the country.
A lingua franca doesn't really needs that to happen : I mean, the historical lingua franca did existed in late Middle-Ages.
But if you meant a standardized language on a national/imperial scale; you'd rather need a mass education (while not mandatory), a centralized cultural policy (such as, for exemple, Richelieu's Académie Française), and mass medias (the importance of newspaper and radios in the XXth on this regard can't be overestimated). Basically, it's about having states enforcing cultural standards one way or another.
A living on Roman state wouldn't be able to pull all of this from nothing, but by virtue having maintained institutions all over a continental scale, would probably slow down the linguistical divergence among provincial elites with themselves would have some influence over their own places. It's, again, more or less what happened with Arab that had to be maintained (for cultural/religious/political reasons) among the various Arabo-Islamic states; and allowed the maintain of an economical and cultural continuum.
It's this kind of continuity that should be searched in order to end up with a Roman language on two levels : institutional Romance (Latin), and provincial Romance, that would be still closely related (would it be only because a living on Roman state would undergo institutional Latin changes on a more global scale).
To wrap this up, maybe it would crush my doubts if we like had examples of vernacular latin in Gaul, Iberia or such places, especially in different periods. I guess that´s a hard thing to do given every text is probably going to have heave Roman and Literary Latin influence.
Well, Oaths of Strasbourg are probably an exemple of vernacular Romance, or at least the very first exemple of first Romance speeches.
Pro Deo amur et pro christian poblo et nostro commun salvament, d'ist di en avant, in quant Deus savir et podir me dunat, si salvarai eo cist meon fradre Karlo et in aiudha et in cadhuna cosa, si cum om per dreit son fradra salvar dift, in o quid il mi altresi fazet, et ab Ludher nul plaid nunquam prindrai, qui meon vol cist meon fradre Karle in damno sit
If it's indeed Old French, it's then quite close to Old Occitan, up to vocabulary. Of course, it's a bit more complex as no medieval Romance text closely follow ONE dialect, but rather mixes them all with one or two dialects or uses dominating the other.
There's some other texts, if as you say, aren't that common at disposal but I don't have access to these texts right now. I could make a medley of what Cantaluse gathered, tough.
That said, I'd point you the Council of Tours of 813 that acts that
latin is no longer clearly understable to people speaking the
lingua romana rustica. Implying, that at some point, it did.