Two thoughts on Berachot 3.
1) The brayta gave three reasons not to enter a ruin: suspicion (of prostitution), collapse, and mazikin (demons). As is his way, Rambam takes the halacha not to pray in a ruin but does not specify the superstitious reason. (He doesn't give any reason.) Yet the gemara outlines three cases (new ruin, two people who are exemplary, who are not so exemplary), and why each reason is necessary. Rambam does not distinguish between these cases.
2) In a brayta:
וְלֹא זוֹ בִּלְבַד אֶלָּא, בְּשָׁעָה שֶׁיִּשְׂרָאֵל נִכְנָסִין לְבָתֵּי כְּנֵסִיּוֹת וּלְבָתֵּי מִדְרָשׁוֹת וְעוֹנִין ״יְהֵא שְׁמֵיהּ הַגָּדוֹל מְבֹורָךְ״, הַקָּדוֹשׁ בָּרוּךְ הוּא מְנַעְנֵעַ רֹאשׁוֹ
See Tosafot and the discussion of how HaGadol as an adjective contradicts Machzor Vitry that Rabba is a verb. Also, about prayers in Hebrew vs. English. However, one thing I learned in Revel was how different Savoraim reworked various masechtot, clarifying the language and sometimes substituting Hebrew for Aramaic. Note the oddity here of HaGadol rather than Rabba, when the preceding word is Shemeih which is Aramaic. See also Masoret Hashas that has that this is girsat Rasha"l, but girsat Rash"a is Shemo, which would be Hebrew. I think there is evidence of a complex editing history, and we can't necessarily derive much from the HaGadol here.
1) The brayta gave three reasons not to enter a ruin: suspicion (of prostitution), collapse, and mazikin (demons). As is his way, Rambam takes the halacha not to pray in a ruin but does not specify the superstitious reason. (He doesn't give any reason.) Yet the gemara outlines three cases (new ruin, two people who are exemplary, who are not so exemplary), and why each reason is necessary. Rambam does not distinguish between these cases.
2) In a brayta:
וְלֹא זוֹ בִּלְבַד אֶלָּא, בְּשָׁעָה שֶׁיִּשְׂרָאֵל נִכְנָסִין לְבָתֵּי כְּנֵסִיּוֹת וּלְבָתֵּי מִדְרָשׁוֹת וְעוֹנִין ״יְהֵא שְׁמֵיהּ הַגָּדוֹל מְבֹורָךְ״, הַקָּדוֹשׁ בָּרוּךְ הוּא מְנַעְנֵעַ רֹאשׁוֹ
See Tosafot and the discussion of how HaGadol as an adjective contradicts Machzor Vitry that Rabba is a verb. Also, about prayers in Hebrew vs. English. However, one thing I learned in Revel was how different Savoraim reworked various masechtot, clarifying the language and sometimes substituting Hebrew for Aramaic. Note the oddity here of HaGadol rather than Rabba, when the preceding word is Shemeih which is Aramaic. See also Masoret Hashas that has that this is girsat Rasha"l, but girsat Rash"a is Shemo, which would be Hebrew. I think there is evidence of a complex editing history, and we can't necessarily derive much from the HaGadol here.