I refer you to:
What Steve Theodore’s answer to Was Latin spoken in the Byzantine empire, even though the official language was Greek? And did Byzantines study Latin texts? said, and what Steve Theodore’s answer to Were the medieval Byzantines familiar with the famous figures of Roman antiquity, like Cato the Elder, Scipio Africanus, or Cincinnatus? said.
In particular, Steve mentions Maximus Planudes’ translations from Latin. Those weren’t a result of a periodic thaw, those were one-offs at the very end of Byzantium. It was only in the last century of Byzantium, with Byzantium reduced to a bystander, the Latins ruling much of the Greek-speaking world, and persistent pressure for Church Union, that Byzantine scholars noticed there was anything worthwhile done in the West at all. It was him, Demetrios Kydones, Prochoros Kydones, Gennadius Scholarius, Manuel Holobolos, George Pachymeres. All 13th century or later. And with the exception of Planudes, who did Ovid, they’re all translating Catholic theology and Boethius, not the Classics.
If people are going to run words together, they don’t so randomly. They run words together when the words form a syntactic grouping. And the stop running words together when they run into a syntactic break.
A clause like “in so far as I am able” is analysed syntactically as:
[in [so far]] [as [I am able]]
There is a break between so far and as I am able. in so far, OTOH, can be argued to hang together as a group. (Even if it doesn’t, it can be argued to be reanalysed into a group that hangs together. Handwaving there, because I don’t care deeply about syntax.)
So there is a natural intuition that prevents you making the leap to insofaras: the as belongs with the following clause, so you can’t run it in with the preceding clause.
As to why the run in to begin with: that’s syntactic reanalysis. in so far as does make sense, if you think about it, but it’s a fairly abstract kind of sense, using a metaphor with spatial extent standing in for validity. Once the metaphor becomes opaque, particularly in a legalese context, people won’t really make sense of in so far word-for-word; so they’ll be tempted to rattle it off as gobbledygook, and thus reanalyse it as a single word.
What they won’t do is extend the gobbledygook to the next word, because they do still understand that as is a conjunction introducing a clause, and not gobbledygook.
Boutros’ Nick, as the villagers called him: Ο Νικολής του Πούτρου. But in the papers of the state and church, he was set down as Nicholas Hadjimarcou. With a c and a dj, because the state was British Cyprus, which acknowledged that the Cypriots pronounces their j’s. The Greek alphabet doesn’t, so Νικόλαος Χατζημάρκου.
Boutros’ Nick married Maria Haralambidou, and they had nine pregnancies, eight births, and seven children who made it to adulthood. George, Helen, Chris, Stavros, Dora, Andrew, and Marc-Anthony. Or at least, Marc-Anthony is what the state knows him as, because the godfather insisted on it. Boutros’ Nick preferred Savva, and Savvakis is what he goes by to this day.
Savvakis is also the only male child who stayed in Cyprus, and the only one who kept Boutros’ Nick’s surname.
The other four boys, George, Chris, Stavros and Andrew, all migrated to Australia, between 1947 and 1970. Chris had switched his surname to his patronymic before leaving; Christodoulos, son of Nicholas, Hadjimarcou (Χριστόδουλος Νικολάου Χατζημάρκου), thus became Chris Nicholas (Χρίστος Νικολάου). Cypriots do this, though I don’t know how frequently.
George came to Australia way before Chris had switched his surname—Chris would have been just 15; but George landed in Sale, Victoria in 1947, at a time when noone particularly felt like deciphering Hadjimarcou. So he found the surname switch expedient. The other two siblings moved two decades later, and they too found the switch expedient—the more so because they all ended up in Launceston, Tasmania, and didn’t want to go by separate surnames.
Now it is the firm custom among Greeks that their firstborn children take their grandparents’ names. Families used to go to war over less: not giving your parent’s name to your child was a grave insult, and that sentiment is only loosening now. It certainly hadn’t loosened in the ’60s and ’70s in the Tasmanian Greek community.
Four Nicholas brothers had four firstborn sons. It was thus an inevitability that there be four Nick Nicholas’s. I’m Stavros’.
And no, my parents thought nothing of it. They were at least merciful enough not to put down my given name in the birth certificate as Nicholas.
Btw. My father went by Nicholas, but he only changed it by deed poll to Nicholas when we were about to leave for Greece. He used Nicholas so he wouldn’t go by a different surname to his brothers; he changed it officially so he wouldn’t go by a different surname to his children.
.i mi ckire do doi filip niuton lenu do cpedu lenu mi spuda .i ku’i pe’i do pujeca djuno ledu’u mi se cmene mu’i dakau