TRS-80 Fragmentation

It is obvious that the TRS-80 line of computers suffered severe fragmentation with differing architectures: TRS-80 Model I, III, and 4(P) are all obviously of a mostly compatible architecture. TRS-80 Model II and 16, 68k based "business" machines TRS-80 CoCo I, II, III (Dragon) TRS-80 PC-x, various rebadged machines from Sharp, Panasonic, or Casio TRS-80 MC-10 (a Matra Alice) TRS-80 Model 100, 102, 200 (rebadged Kyoceras) So, obviously there were several good sellers in there, and of course for every good seller there's at least one bad seller. The PC line were mostly replacements for calculators that were programmable, and the Model 100 derivatives were mostly used as appliances rather than general purpose machines. Aside from that, it seems like Tandy more than most went off in the weeds with their own wide variety of machines instead of settling on a common architecture. Do you think that if they had, say, revised and extended the Model I system to color/80 column that the rest would have been mostly redundant?