I don't believe that AoOA thresholds have been very well documented at this point, so I don't have an answer for you. I would definitely like to know though if you do see a score improvement once hitting 48 months, so please stop back in and follow up!
It's a pleasant surprise when I hear someone say this. People here are constantly being endorsed for providing "data" without any investigation into its quality. The belief apparently is that the more "datapoints" we have the more likely we are to reliably know the answer to something. But in fact 100 datapoints with no careful look into quality is just 100 x 0 -- we are no closer to knowledge than we were before.
Thanks for your restraint and awareness of the problem of confounding factors!
PS. Contributor BBS is especially aware of the problem of CF and is good at going the extra mile to ask a lot of followup questions.