I guess the best way to respond to this is that too much of the policy-making associated with lawful access has not been evidence-based. It's based on the occasional anecdote without strong evidence.
It seems to me that if we are in the realm of trying to ask what the appropriate amount is—a year is clearly way too long—what we need is far more evidence, frankly, about when there are instances when that metadata is actually needed and how long it usually takes before you might be able to get that order.