Here Oliver Cowdery originally wrote “the remainder was” in both manuscripts. Both instances of was were corrected by crossing out the was and supralinearly inserting were. In the original manuscript, the correction is written with slightly heavier and uneven ink flow, which suggests that the correction in 𝓞 could be secondary. In the printer’s manuscript, there is no change in the ink flow at all, which suggests that the correction in 𝓟 was virtually immediate. There are two possibilities here: (1) Oliver consciously decided to correct was to were as he was copying from 𝓞 into 𝓟, with the result that he made the change virtually immediately in 𝓟, then turned to 𝓞 to make the same change; (2) Oliver independently wrote was in both manuscripts, correcting it somewhat later in 𝓞 (perhaps when he read the text back to Joseph Smith) but virtually immediately in 𝓟.
Under Mosiah 10:14, I argued that evidence throughout the manuscript shows that Oliver Cowdery sometimes accidentally wrote was in place of were but that there was not much evidence that he consciously emended was to were. In that discussion, I missed noting the difference in ink flow for the passage here in Alma 46:33. Consequently, there are at least two places where one could argue for conscious editing of was to were by Oliver, in Alma 43:47 and here in Alma 46:33. Nonetheless, there are numerous instances in the earliest text of the singular was being used with plural subjects, obvious instances of nonstandard subject-verb agreement that Oliver never corrected (see the examples listed under 1 Nephi 4:4).
Here in Alma 46:33, one could propose that Oliver Cowdery decided to emend the was to were because of the were in the conjoined predicate that follows: “and were taken back into the land of Zarahemla”. But elsewhere in the text, there are instances of conjoined predicates with mixed use of was and were, yet these were not corrected by Oliver (only later by Joseph Smith in his editing for the 1837 edition):
In all these instances, Oliver was the scribe in 𝓟.
The use of the were for the noun remainder is consistent with usage elsewhere in the text, even when there is no intervening of-phrase (as in the first example listed below):
Oliver Cowdery’s tendency to write “the remainder was” in Alma 46:33 for both manuscripts may have been due to the fact that remainder is singular in form but plural in meaning (providing we are dealing with objects that can be counted). Nonetheless, he did not make this error when he wrote down Alma 2:11 in 𝓟 (𝓞 is not extant there): “ the remainder were called Nephites”.
Summary: Maintain in Alma 46:33 the use of the plural were with remainder, the corrected reading in both 𝓞 and 𝓟; the use of “the remainder were” is characteristic of the Book of Mormon text.