Investigate whether slice_push_back and similar methods mutate arrays directly

[jfecher]

Link to the original PR: #6685 (comment)

Need to investigate for make_array deduplication whether we also shouldn't deduplicate make_array instructions which create slices which are later used in slice_push_back and similar methods. If these methods mutate the slice directly, and a later slice is deduplicated to be the original, then the later slice may have the incorrect mutated value instead of the original.

[jfecher]

Each slice method (currently at least) copies the slice and returns a new one, so the check in the original PR should be fine.