As I’ve said many times, it’s not a time bomb, and clearly if something passes assay criteria after 3 years of stability (that’s generally 100% of label +/- 2.0%, IIRC?) then there’s lots of reason to believe that it’s going to stay within that range for considerable time more. On the other hand, I have tested things with impurity criteria that were something like no more than 0.5% of some degradation product, and on day 1 it’s at 0.001% and after 3 years it’s at 0.35% or somesuch. How to predict when it hits that 0.5% mark? Just like you know your milk is still drinkable a day or two after it’s “best before” or expiry date, but there’s nothing to say exactly when it will turn bad, you know?
The stability protocols don’t require more than what is currently done, and these things are pretty much internationally agreed upon. If the FDA, TPD, etc want to change and extend the requirements in order to have longer expiration dates and longer shelf lives, they are free to do so (provided the stability data supports it), but then we’d simply be asking if the drugs are safe after 5 years instead of after 3. Consumers will still question it, without thinking about where the date on the package actually comes from.
Of course manufacturers want to make a profit, of course they want to have a drug on the shelves for less time so they can sell more…but also, of course, they want that drug to stay safe over it’s shelf life, because the fallout from a recall can be damaging. And so the dates on the packages reflect both the balance of safety and the balance of profit.
The point, though, is that stability past what is currently required by regulatory bodies simply isn’t being tested. Why would a manufacturer test more than is required by law? No matter how long the FDA changed these requirements to, the fact is there would still be no data whatsoever to support stability beyond the end of the required protocol length. Aspirin could be proven to be stable for 30 years…there would be no data to support whether it’s stable for 31.
And your own cite mentions testing of the military stockpiles, and says that while “many” of the products were stable longer than labelled, many were not, even when stored under optimal conditions, which pretty much never happens in pharmacies or patient homes. Temperature variations, humidity variations, etc will cause unpredictable effects on stability and degradation.
So at the end of all this, you’re back where you started. Some drugs could have a longer shelf life, some might not, and that data is collected according to internationally recognized regulations. A day after that, though, no data exists, and someone will still be asking this question on a message board.