... just some of the differences:
- the startup procedure is different, with the "V" it waits for autobauding for some time (this can be inconvenient if you want to start your application immediately, and detrimental, if there is something transmitting continuously to the mcu even during it is reset, so it could possibly "see" the "U" character for autobaud and stays in the bootloader rather than starting the application) - the good news is that you can modify the bootloader, the bad news is that there is no alternative bootloader around...
- the bootloader "commands" are different
- the IAP "commands" (and calling procedure) is different
- the "V" is erased in 128 byte blocks, while the "C" is erased in "big" (several kB) blocks
- for "V", the bootloader+ISP is in a 8kB shadow flash starting at 0, while for "C" it is the topmost 1kB (in fact, 2kB, as we learned here recently from Bill Houghton, but 1kB from it is normally "hidden")
- the 6/12 clock can be switched "on the fly" by software for "C", but is fixed at programming time at the "V" (and even less conveniently, once you program it, you need a parallel programmer to erase it)
- there is an extra SPI port at "V"
- they behave differently when you do PUSH SP/POP SP :-)

Jan Waclawek