The Gemara in Makkos 23b states:
דרש רבי שמלאי שש מאות ושלש עשרה מצות נאמרו לו למשה שלש מאות וששים וחמש לאוין כמנין ימות החמה ומאתים וארבעים ושמונה עשה כנגד איבריו של אדם
R. Simlai when preaching said: Six hundred and thirteen precepts were communicated to Moses three hundred and sixty-five negative precepts corresponding to the number of solar days [in the year] and two hundred and forty-eight positive precepts, corresponding to the number of the members of man's body. (Soncino translation)
However, R. Moshe ben Avraham of Przemyśl (died 1606) in the introduction to Mateh Moshe reverses this and states that there are 365 positive precepts and 248 negative precepts.
כי ברא אותו מחובר ברמ"ח אברים ושס"ה גידין וכנגדן צוה צורינו שס"ה עשין ורמ"ח לאוין להורות כי לא נברא האדם רק לשמור ולעשות ולקיים את כל הדברים האלה
I would think that this is simply a mistake; however it is a particularly egregious mistake for a renowned rabbinic figure to make. Most tinokos shel beis rabban know that there are 365 negative precepts and 248 positive precepts.
Additionally, a mere few pages later R. Moshe quotes the Gemara in Makkos, but he leaves out the part which specifies that 365 refers to the negative precepts and 248 refers to the positive precepts.
חש בכל גופו אמר זה על דרך דאיתא במסכת מכות דרש רבי שמלאי תרי"ג מצות נאמרו לו למשה בסיני שס"ה כמנין ימות החמה רמ"ח כמנין אבריו של אדם
Is it purely coincidental that he left out the very part of the citation that he reversed earlier?
On the other hand R. Moshe also wrote a book of the 613 mitzvos (Sefer Shel Taryag Mitzvos), and he does not seem to have an entirely different count (as would be necessary to reverse the numbers of positive and negative commandments).
Is there any possible way to interpret R. Moshe's first statement without simply rejecting it as a mistake? Or are there at least any subsequent rabbinic figures that noted this glaring mistake?
Of course, it is certainly possible that it is just a typo. However, all five editions that I checked (Cracow 1591; Frankfurt 1720; Warsaw 1876; London 1958; Jerusalem 2011) have the same error (which doesn't necessarily mean anything, as they might have just copied from each other) and none of them note this incorrect wording (including the newest edition that has extensive footnotes). Moreover, this would not help explain why the later citation of Makkos left out the part that 365 is for the negative precepts and 248 is for the positive precepts, nor would it explain the fact that the error is not noted in subsequent rabbinic literature.