First of all, great work guys! Love the additional USB ports, I/O, proper mounting holes, better power supply.
Regarding the Videocore causing limitations for I2C-0
The ID eeproms would also cause issues if the Videocore truly used it randomly.
Logic dictates that the Videocore uses a range of I2C addresses that do not conflict with the EEPROM's addresses.
The Foundation should document what I2C addresses are off-limits due to Videocore, which would tell us what I2C addresses are safe for non-EEPROM peripherals. Much better than allowing for random conflicts - and there should be no NDA issues, as the Foundation would not have to document the *function* of the I2C addresses used by the Videocore.
One of the I2C buses need to be "reserved" for use by Videocore to talk to the camera (and probably Display as well). Use of I2C0 on the ARM is not recommended as e.g. if you fire up Raspicam then Videocore will assume it has exclusive control of the interface, and start talking to it.
There is no electrical reason why you cannot connect random things to I2C-0, but unintended operation (including crashes/lockups) may occur if/when both Videocore and the ARM try to talk to the same piece of hardware at the same time.