In teaching and learning Masechechet Shabbat I've [so far] run into 3 or 4 difficulties with both Rif and Rambam
Here is my working list -
1. The famous case re: hatmanah at the end of Chapter 2 on the Mishnah of "tomnin et hachamim" bein hasmashot [lit. S'feiq hasheicah s'feiq einah chaseichah]
The difficulties here are well documented as early as Ro"sh and B"Y. More later BE"H
2. The missing insertion in Aramaic in a Braitta at the end of Chapter 4. This explanatory comment in Aramaic is clearly not native to this Braitta. The issue is when did it get there? And Why do Rif and Rambam SEEM to ignore it?
3. TB Shabbat 6b Abbayei makes a Hilluq between deserts being r'shut harabbim or a carmelit to [dialectically] address the Talmud's conflict between two braittot. Neither Rif nor Rambam have this Hiluq. Furthermore Rambam takes the puzzling position that Midbar [at least by default] is a R'shut Harrabim. That issue is also important, but not as critical to me as - "How come Rambam does NOT have Abbaye's differentiation at all?" For that matter Rif does not either, though it's conceivable that he may have seen no need to have it.
The common denominator leads me to suspect a "generality" a Klal.
Hypothesis: Rif and Rambam had a poor quality edition of Masechet Shabbat
Ergo many difficulties can be answered by attributing this at the root cause.
There are of course other possibilities. I simply intuited this when researching the Rambam by using the Rif. The Kessef Mishnah advocates using the Rif to ferret out the Rambam's methodology. It usually works.
I will BE"H outline my findings on each case. AFAIK there are other approaches to addressing these issues but no hard evidence that Rif/Rambam had a girsa/text like ours.