The one thing that I am afraid woul absolutely fail is that it wouldn't be consistent enough for great accuracy. Ball powder will do one thing well, within microseconds, that entire charge will be in full ignition, so, overall, every round should be practically identical. Using random, chaotic events. Flip 1,000h coins, and no matter how often you do it, you're going to come up very even.
Making the actual solid core, especially multi stage, if there is a hiccup of any sort,the acceleration falters, you may wind up with high standard deviations in MV. Another reason why that ideA is probably an orphan cause unless experimentation shows that mil spec 5.56 velocities could be boosted at least 2-300 fps,or some other huge benefit comes up.
If the program ever gets off ground, they will have to create a completely different cartridge, as it would probably not be safe in standard arms. Higher pressure at gas ports and muzzle, and having a very high pressure blast at muzzle may not be good.
Just something I've thought about for many years, but never expect to see.