I have been learning how to do paleographical work on Latin Medieval manuscripts since August. If we are taking Ciceronian Latin as the standard, every Medieval manuscript I have looked at so far has diverged from that standard in spelling and grammar.
In many cases, the scribe seems simply to have made mistakes. In other cases, the spelling of a word, or its grammatical use, changed throughout the centuries, so in those cases, "correct" or "incorrect" is not really the right word-it is simply different from Classical Latin.
The issue is complicated because in most of these manuscripts, many words are abbreviated, and in some other places, the handwriting is unclear. So sometimes what words are actually there, and their spelling, comes down to the Paleographer's interpretation. When the words in a manuscript are unclear, one's inclination will be to use correct grammar and spelling when filling in the gaps. But was that the spelling and grammar which the scribe intended? That's not an easy question to answer.