[JesseK]

Just a small anecdote from coaching on field: historically, entanglement blame has gone to the robot that has the offending mechanism, not the robot that interacted with it. For example, if BotA has an arm that gets entangled on the frame of BotB, it is BotA's fault for having the arm sticking out rather than BotB's fault for driving into it. If the arm gets entangled in an extended appendage of BotB, the faults offset.

I agree with Al that the GDC will probably not answer the question since it directly requests approval of a specific design. However, as stated, this is a risky design. Consider the implications of another robot pushing on the flexible point with their drive train frame (not an arm or anything else). Or, consider another robot driving over any cable in the middle (not all robots are typical 18" box drive trains). If it's too high a risk yet you still want the balancing flexibility, perhaps consider alternative derivatives to the strategy: does your robot need to be separated throughout the entire match?

Good design adage: design your robot to withstand abuse by the environment rather than expecting the environment to conform to your design.