I have some doubts about the agreement of verbs. Two examples from Adler’s exercises in his Practical Grammar follow:
As for Example 1: Why fecisti instead of fecerunt? This Latin sentence is constructed like “Non tu mihi dolorem fecisti, sed liberi tui.” But why? Why not let “hurt” agree with the closer “your children”, which would result in something like this: “Non tu, sed liberi tui mihi dolorem fecerunt.” (I admit that I may be influenced by the German language in this case).
As for Example 2: Why plural incusant instead of singular incusat? It seems to be an example for Adler’s Rule B, according to which “after a collective noun the predicate is either in the singular or the plural.” Now, there is no well-known collective noun in the sentence “Francis complains of Eleanor, and Eleanor of Francis”, but Adler says that this rule is also valid for “uterque, quisque, alter…alterum, etc.” Still, to me it sounds weird. Isn’t it more natural to say “Franciscus Leonoram [“incusat” implied but not expressed], et Leonora Franciscum incusat?” The same problem arises in the fourth sentence of this example when Adler uses volunt instead of vult.
The first one seemed odd to me as well with fecisti, but I think that is me reading English syntax into it. non tu mihi dolorem fecisti, sed liberi tui wouldn’t bother me, and the only difference is the word order, which tells me that my native English word-order-based-syntax is why I don’t like it. Hopefully mwh will chime in because I’m not sure about it.
The second construction I know is correct. English wants a singular verb in such a construction, seeing it as two clauses with ellipsis of the verb (John runs on Mondays, Jim runs on Tuesdays), whereas Latin tends to see instances like this as a compound subject in one clause with a plural verb (John on Mondays and Jim on Tuesdays they run). That is unacceptable in English, but not in Latin.
You don’t want mwh to chime in on the first one, you want a real latinist, which I don’t pretend to be. But for what it’s worth—very little, and I sincerely mean that—I think fecisti is at any rate preferable. It’s the tu that carries the weight, and it’s that that needs completion, non tu … fecisti directly answering the effecin’ ego question (the simplex representing the preceding compound as often); it’s as if sed liberi tui is brought forward into the non tu … fecisti clause, not exactly parenthetical but elaborating the principal statement.
In 2, I don’t think the singular would be wrong, but it would lessen the reciprocality. If the verb came in the first half (after Leonoram, and after Francisci) it would of course be singular; coming at the end as it does the plural makes a single sentence, as calvinist says, rather than a double one with repetition of the verb “understood.”
But this is unsupported ipse dixi, and as such perilously unreliable.