WR Moore has a good memory. The Harvey 2-piece bullet goes back to the 1960's or earlier, another attempt to keep leading at a minimum in higher velocity loads while also providing a softer point for terminal performance.
Harvey also designed a bullet mould that allows a zinc washer to be cast into the base, supposed to act somewhat like a gas check.
I think what the OP is refering to (2-piece construction bullet for penetration) is a project undertaken for the FBI back in the 1970's. The subject matter was to improve penetration through automobile glass, specifically windshields (very tough stuff indeed, and the curvature of many windshields makes for a lot of potential angles of impact).
At just about any angle, very few handgun bullets will reliably penetrate through windshields without deflecting in path considerably, making it very difficult to reliably hit a target inside an automobile. At angles approaching 45 degrees many handgun bullets will not penetrate at all, being deflected off at unpredictable angles.
The result, as I recall, was dubbed the KTW Penetrator, featuring a truncated cone shape with flat point and very sharp edges, and having a base or insert (don't remember which) of a dense and hardened material (tungsten carbide perhaps?) intended to punch through in the event of impact causing the bullet nose to deform. Testing showed only a marginal improvement in penetrating automobile glass, and the project never resulted in much interest, although reported in the magazines of the day.
This caught my eye as a young copper, and I did a little experimentation of my own on several junk cars, using .357, .45ACP, and 12-gauge (00 buck and rifled slugs). At any significant angle whatsoever, I found that most shots were completely deflected by a windshield, and those that actually penetrated never continued on the original path. So, hitting an intended target inside a vehicle is actually quite difficult to do, even at close ranges.