I wouldn't be at all surprised to discover string-hop or jump if you prefer, in all locks that fall away below the top of the stock. With medieval roller-nuts you can keep the hop down to a minimum by making sure the nut is lightweight and rolls very fast. (and that's why another maker referenced somewhere above, resorted to drilling holes into his bronze roller-nut... I did the same thing on my first bow) Making the lugs slightly hooked on the rear will also delay hop a micro-second or so, and every little bit helps. If you have good contact between the bowstring and butt of bolt, whatever hop remains should be no problem, but I expect it's still present... which could be one reason that modern target bows usually push the bolt with a slider, rather than letting the string do it (the slider also assures the shooter than he has pulled both limbs evenly) Locks that release from the top can rely on the bolt-track to prevent the string from going anywhere, so no string-hop, though I suppose there could be down-forces that serve to flatten the string. In some cases, you might even get the string UNDER the butt of the bolt for entertaining results.
High-speed photos I have seen suggest substantial oscillation in the string on release isn't at all uncommon. Certainly crossbow strings with a hidden broken strand can precipitate misfires even though they appear normal otherwise.
A substantial cowl or string-snubber should prevent hop from being a problem on most locks mounted in the stock. I would include them as a matter of course on any modern design. Medieval locks generally don't feature them and since that's what I make...