That last issue can be handled by encoding each logical bit as a pair of physical signals (like low/high for 0, high/low for 1). I don't think there's any reaso

That last issue can be handled by encoding each logical bit as a pair of physical signals (like low/high for 0, high/low for 1). I don't think there's any reason you can't implement your reversible counter in binary.