Without going into too much detail, rather than “numerous” sahabahs with perfect recollections of the Koran, there is compelling evidence that these earliest sahabahs had different and differentially complete memories. How else is one to make sense of the ahadith (repeated in one form or another at least seven times) concerning the last verse being found in the memory of only a single man; Abi Khuzaima Al-Ansari? Doesn’t that require the understanding that every other sahabah had an incomplete memory?
How does one further account for the conflicts among those sahabahs that required the Uthmanic rescension?
In fact, how does one account for Uthman’s rescension at all?
It is pretty clear that the “perfect preservation” of the Koran did not commence until after the rescension. And even then, we could get into a wonderful discussion of the multiple readings, but why bother. The point is made. Second, giving the fact we have a historical record of the event during which the Koran was standardized and competing versions burned, the maintenance of a standard since that time hardly qualifies as suggesting that Mohammed would have even recognized the writings. The completely human engine for that standard is evident and obvious. We have in our possession, at best, the musshaf of Uthman. We really do not know what the musshaf of Muhammad contained, and how different the two might be.
the readings (القراءات) is simply readings!! the fact that there are many readings of quran is related to the history of the arabic language.
at the time of Messenger, arabs were speaking arabic with different (tongues). (a normal phenomena in all languages, such as there is different tongues of french in france).
there were slight variations in the pronouncement of words among arabic tribes. It is also reasoned that arabic language wasn't standardized yet. so the same word might be written with different spellings.
It wasn't until a century later, that muslims established grammar rules for the language and dictionaries of the words of the language were written with the spelling of words fixed.
returning to the period of the Messenger,
The Messenger was from Quraish tribe, which has its own tongue. Quran was downed from the heaven with this tongue.
To ease the recitation and reading of quran to the rest of arabs, an allowance regarding pronunciation/some grammatical variations/... was given. it was reported by the messenger himself that there are (أحرف سبعة) seven variations in the reading scheme.
This fact was known and accepted by sahabas and later generations of muslims.
By the adoption of Osman's mushaf, the spelling and drawing of words were fixed. any other spelling were abandoned in the sake of unification. but the allowance for the different readings remained provided that it is compatible with the Osmanic musshaf.
in the sake of unification, 10 readings were selected and fixed. no other variations were allowed. These readings were confidentially traced to the Messenger. each reading has its chain of (tellers) that each one of them heared this reading from the mouth of the next one in the chain up to the Messenger. (of course this was backed by the written musshafs)
The reading includes fine details on how to pronounce each word, how long to pronounce each letter (for example the two forms wood and woooooood were distinguished) and other many fine details. All these readings are rooted to the Messenger.
The islamic literature included so much detail on the description of Quran words, pronounciation, detailed reading instructions.
the claim that different readings are different versions of quran is like saying that the two statements
( it is a cat) and ( it's a cat) are different versions.