> Spelling is conventional and spelling and alphabet changes don't necessarily correspond to anything meaningful in the spoken language
On the contrary, spelling is highly idiosyncratic until the 18th century, and until then it was tightly correlated to the sounds of spoken language. Shakespeare didn't even HIMSELF have one way of spelling his own last name. That's how non-conventional spelling was until pretty recently.
You can even see it in these examples, words like "maiſter" in IIRC the 1300s example. Which becomes "master" later in English, but remains Mäster in Frisian (the closest Germanic language to English) and is also mäster in Swedish.
I think you are missing my point. Just because spelling can be inconsistent doesn't mean it's not conventional. We agree that certain letters and combinations of letters correspond to certain sounds--that's a convention. We could just as easily remap the letters in our alphabet to entirely different sounds from the ones they represent today and the resulting written text would be, on the surface, entirely incomprehensible, because we no longer understand the conventions being used.
In this particular case, there are several glyphs used in the older texts which we don't use any more today, which makes the older text both appear more "different" and, for most people, harder to read. But this is an artificial source of difficulty in this case. I acknowledge your point that some other spelling differences track pronunciation differences but this isn't always true.
As far as pronunciation changes that aren't captured in spelling changes, this is true most obviously for a lot of words whose spelling standardized during or before the Great Vowel Shift, like "day".