I have done a fair bit of testing related to heat treatment. Started several years ago helping my son with a science fair project about heat treating metals. We used lead since I had a bunch and it is relatively easy to cast samples and apply heat treatments.
We were casting with isotope lead that had been tested by XFR and was 96% lead, 3% antimony, 1% tin. Cast a bunch of samples and tested hardness using a Lee hardness testing kit. Initial results are here
About a year later my daughter did a follow-up project where we retested all the samples after 1 year. The short answer is the untreated bullets trended towards an 11 bhn, and the treated bullets trended towards a 23 bhn.
I shoot heat treated bullets in both handgun - 38/357 magnum, 10mm, 44 magnum and 45 acp, and rifle - 6.5 carcano, 308, 30-06, 300 win mag. I also cast for a friend who shoots them in 30 carbine, 30-30 and recently 6.8 SPC (still to be tested).
In general, I find heat treated bullets to be too brittle for much more than paper punching unless you boost the tin content up above 2%. With equal/near equal tin/antimony content the bullets are hard, yet malleable enough to hold together when encountering bone, cartilage and other anomolies that accompany actual anatomy.
One interesting note I would offer. It has long been held that oven heat treatment after the fact is superior to quenching from the mold in three areas: higher peak hardness, greater hardness consistency, and better surface properties as the lead is not work softened after hardening.
I could not say that my childrens testing supported the first two, and the testing provided no data related to the third. The hardest bullets we cast were quenched from the mold (bhn 32), not from an oven (bhn 25). There was no difference in deviation or extreme spread between batches that were quenched from the mold versus from the oven.
I will say that during casting we were meticulous about casting cadence, and benefited from a mold that released cleanly with little effort. I can agree that if a uniform cadence is not used, or if a mold has "sticky" cavities that delay the release of the bullet by a few seconds there may be more deviation. Temperature at quench time has a huge impact on hardness. This is the primary reason I believe the quenched from the mold batch reached such high BHN values. Those batches were cut with the sprue barely solid and dropped smartly from the mold.