Several good points have been made regarding what is really important in a regulator's performance. One thing I noted when working as a tech is that some brands of regulators show much more improvement before and after annual service than others - or looking at it the other way, some brands of regulators experience more degradation in performance over the average diving season. This became pretty obvious working in a small shop on the same regulators year after year where you knew how it performed when it went out and how it performed when it came back in for service the next year. Some of it is due to design but most of it is due to materials selection for the soft parts.
As an aside one the companies where I notice the most increase in performance with an annual service also allows for a two year service interval with only an inspection required every 12 months. So in that case the tech has the not so wonderful choice of sending the reg back out for another year in safe but less than peak condition or making the customer potentially mad by suggesting a full annual service - one of those cases where before and after performance numbers can be very valuable.
So in effect, I think there is more value in testing a reg with several months and 50 or so real world dives on it if you want to determine how a regulator will perform in real world conditions.
I also agree with Phil that testing a super well tuned reg can be misleading as that same setting may not provide and adequate degree of reliability as the reg is used over the season - another good reason to test regs that are middle aged in terms of their annual service schedules.
As an aside one the companies where I notice the most increase in performance with an annual service also allows for a two year service interval with only an inspection required every 12 months. So in that case the tech has the not so wonderful choice of sending the reg back out for another year in safe but less than peak condition or making the customer potentially mad by suggesting a full annual service - one of those cases where before and after performance numbers can be very valuable.
So in effect, I think there is more value in testing a reg with several months and 50 or so real world dives on it if you want to determine how a regulator will perform in real world conditions.
I also agree with Phil that testing a super well tuned reg can be misleading as that same setting may not provide and adequate degree of reliability as the reg is used over the season - another good reason to test regs that are middle aged in terms of their annual service schedules.