I'm hopeful it will be warm enough on Thursday to Whink some chips.
I torched the TMS1400 from Split Second today and took pics: www.seanriddle.com/tms1400.html
I also uploaded a 6000x5000 pixel XCF: www.seanriddle.com/splitsecond.xcf
It has double the ROM space of the TMS-1100; the ROM array is twice as wide. I haven't figured out the mapping yet, but my guess is that the rows are mapped the same as the TMS1100, the first 16 bits in each row are bit 7 for the 16 pages in chapter 0 (in the top half, chapter 1 in the bottom half), and the second 16 bits in each row are bit 7 for the 16 pages of chapter 2 (in the top half, chapter 3 in the bottom half). The next 16 bits would be bit 6 for chapter 0, then bit 6 for chapter 2, etc.
The O output PLA looks the same, but with the full 32 possibilities instead of the 20 that the TMS1100 was limited to.
The instruction PLA looks the same but rotated 90 degrees counter-clockwise on the die.
The only datasheet I've found for it is on Kevtris' site, linked from my page above. To handle the additional ROM chapters, the COMC instruction was changed to TPC. It also has a 3-level subroutine stack!
The rest of the Split Second hardware is pretty simple: 6 buttons, a speaker, and 53 LEDs in a 7x8 matrix.