In some ways risk is like an air filled balloon. Early on, it's pretty easy to reduce risk (use your hands to squeeze the ballon to reduce its size). Over time it gets harder to reduce the total risk further. Squeezing the balloon in one dimension results in it bulging out in a different direction.
Sometimes this increased risk elsewhere is anticipated. Sometimes it is not - until it happens. It becomes a trade off, and you can never get it down to zero.
Mh370 may have come about partly through our efforts to mitigate 9/11.
So this question is a balance of risks. On balance, I wonder if the risk of being able to turn a transponder off is greater than the risk of not being able to do so.
I accept that a functioning transponder may not have saved lives, either here or 9/11. We will never know for sure.
But we would know what had happened in real time. With 9/11 it might have permitted the shooting down of UA93 had passengers not taken things into their own hands. A hell of a decision, but what would you have authorities do if it was headed towards a major population centre. We would likely have the recorders for MH370. And the sure knowledge you couldn't disappear may act as a deterrent.
As the world moves into a period of higher risk, the (acknowledged) risk of not being able to turn off an electrical item may actually now be the smaller risk.