Ktzos 28:1 - chiyuv la'tzeis y'dei shamayim (II)

Last week I started a Ktzos and did not finish off the topic. See here for part I: the Rishonim ask what is the difference between a case if ba bamachteres, where Rava apparently did not wish to take his property back from thieves because they had a chiyuv misa and we apply kam lei b’derabbah minei, and a case of esnan zonah with a close relative, where despite the chiyuv misa, the financial obligation to the zonah still exists as a chiyuv latzeis y'dei shamayim and causes the korban to become assur.
Tosfos in Sanhedrin writes that Rava also agrees in the case of ba bamachteret that the thieves should make restitution latzeis y’dei shamayim. However, he still refused their money. These thieves, Tosfos explained, turned over the money only because they thought they were chayav m’dina, which is not true. Rava did not want to accept payment under false pretenses. (Interesting: although they were obligated to make payment anyway, accepting money under false pretense is still assur!) The Ran (B.M 91) answers that the thieves may indeed have not been required to make payment. The case of esnan zonah is fundamentally different from the case of ba bamachteret because the obligation to the zonah is a contractual obligation, and a contractual obligation is not cancelled by kam lei b’derabbah minei, only a financial penalty is cancelled.
What is the point of disagreement – why would Tosfos not accept this sevara? You could perhaps say the issue is how to treat a contract – is keeping an agreement a natural law obligation that doesn't need the court's intereference and hence remains untouched by considerations of kam lei b’derrrabah minei (Ran), or is a contract only meaningful because it is enforceable by the courts, and where the court is blocked from interfering because of kam lei, all bets are off (Tosfos). But AddeRabbi’s nice lomdus in a comment to distinguish between the ba bamachteret and esnan cases got me thinking further. He wrote:

getting fancier, even acc. to rava, we can ask, does kim lei erase the second crime, or the second punishment. in certain cases, it's obvious that it's the latter. for example, your case of etnan. there's one crime with two potential punishments, so the latter punishment is not enforced, but that doesn't mean it doesn't exist on some moral plane… that's the definition of chiyuv latzet yedei shamayim - you have a very real obligation, which simply has no enforcement mechanism. with regard to ba bamachteret, there are 2 actions - breaking in and stealing. the forfeiture of life that comes with breaking in makes us focus on that and ignore the theft. we ignore the second crime itself. stated differently, the context of machteret relegates the theft to trivial status. that's what 'be-damim knenhu' means. in that case, there's no chiyuv latzet yedei shamayim even because no debt was ever accrued (and again, chiyuv latzet isn't equivalent to 'moral obligation').

I want to borrow part of this to explain Tosfos. In the case of esnan, or shor b’disho (B.M. 91), the obligation to pay results from a specific act having occurred. Kam lei b’derabbah minei perhaps does not just mean that the financial penalty that results from an act is cancelled (Ran), but it means that in the world of dinei mamonos it is as if the act never occurred. Even if the obligation for payment is a result of normal dinei mamomos and not a penalty, the trigger that caused the financial obligation to occur is missing, and all that is left is a chiyuv latzeis y’dei shamayim.