I'm pretty confident that a MM47 emulation core is possible. But it'll have to go to the endless virtual TODO list for now. If it's similar enough to the COP400, it can be added after someone(can be me) refactors our cop400 core.
I pulled the MM4799 die off the Mattel Basketball PCB and took pics of it. This is the game that someone had wired the 9v battery clip backwards, and that vaporized some of the metal on the die and some of the bond wires, which sputtered onto the die, obscuring some stuff. But most of the die is clear, and I'm hopeful that using some Whink on it will clean it up.
It has the same amount of ROM and RAM and it looks like the same pinout as the MM5799 described in the NSC 1977 MOS LSI databook on bitsavers.org. The ROM array is laid out oddly- it's 1.5K, so 24 pages of 64 bytes. But instead of having a column for each of the 24 pages, it has just 16 columns per bit, and instead of 64 rows, it has 96. So the 8 other pages are in the extra 32 rows. I'd like to figure out the bit arrangement and see if disassembling the ROM contents as MM5799 opcodes results in good code.
I don't think it's similar enough to the COP400 to be included in that driver, but I could be wrong.