I can't contribute anything to the question of calculating hardness by weight, but I found the LASC links very interesting. I was very impressed by the link, under Cast Bullet Notes, to the hardness checker test, where different people measured a batch of bullets and got remarkably different results using various checkers. This post did motivate me to try something I've been considering for some time now. An auto body tool was used to drive a .061 decapping pin into various metals, with results shown below. I'm sure this has been done, and there may be methodological flaws, but as an LASC post noted, the actual BNH number is irrelevant as long as it gives you useful information, and the results I got are very consistent.
The pin had to be held in an expander button, and I cut the stem off anticipating use in a 22-250, but this proved inadequate, so a 25-06 button was used. Unfortunately, the longer length prevented the button from fitting properly in the tool, so a certain amount of mis-alignment was unavoidable. Also, ingots tend to be uneven and a crooked strike raised a dimple of metal around the hole, complicating measurement.
Of the material checked, the pure lead and linotype alloy were purchased online and their purity can not be verified. The wheel weight ingots are a blend of internet metal and the last of my stock, secured from a tire shop about 25 years ago. I behaves about as I would expect it to. The 50/50 lead-tin is from Rotometals, a trusted source. H and VH are two of my blends.
I plotted results for known BNH metals on a graph, and it suggests the relationship is algorithmic rather than linear, as the SAECO data indicates. For your consideration and possible amusement, I offer:
