Isn't any method of synchronizing clocks likely to be vulnerable to the same sort of counter resynchronization issue that this attack exploits?

If you can do that, you probably have two-way communication, and this issue should be moot.