I work for a company that uses FAID as part of it's FRMS. Basically they have a set of rostering protocols that are loosely based on their old exemption to CAO48 and then the FAID score runs along side it as a backup. I will start by saying that despite the flaws in the system itself, i think my company does a fair job in administering it and I've never personally had a negative experience from a fatigue report.
I think FAID is a useful tool but it should not be used on its own. For one, FAID only looks seven days back. The theory is that fatigue due to lack of sleep is only cumulative over a short period of time, but we all know the reality is that working hard over a long period of time leads to long term fatigue. I believe this is because long work hours leads to an increased need for rest and therefore an increased chance that you won't get the required rest in the time available. FAID does not account for this at all. So if you used FAID and nothing else then you could roster someone 365 days/year.
FAID is also based SOLELY on the amount of time you have available for sleep as determined by your duty periods. For example if you sign off at 4pm on day one and sign back on at 8am on day 2 it will consider you to have had the opportunity for a full night's sleep. What it does not do is consider your working environment and how fatiguing the actual work is.
I believe CASA dropped the ball on this to some extent, they don't want FRMSs to include any limits on flying hours, they want it to all be based on how much duty you're doing. This completely ignores the effect that flying has on fatigue. The whole thing is designed around how much sleep you are getting and ignoring the fatiguing effects of your actual work. Now I sleep a hell of a lot better after a hard day of work, so the work itself obviously has a significant effect on the body and its need for rest.
Another place where the system is screwed is that it ultimately relies on pilots to call in fatigued if they are not fit to work. That sounds great in theory but it's not good in practice and should be avoided, that is, the rostering system should be conservative enough that it is very very rare for a pilot to have to report fatigued.
We know that fatigue affects decision making, and you're relying on the fatigued person to be the one who makes the decision, not good. It relies on the pilot to not be influenced by any external factors. There are lots of things they might stop a pilot from calling fatigue when they should. Such as being on the last day of an away trip and wanting to get home, or being new to a company and not wanting to rock the boat (perceived pressure), or just believing that, despite assurances to the contrary, there are negative consequences to calling in fatigued.
Ultimately a pilot will have to be the one who has the last say, because no matter how conservative the FRMS they work to, there will be times when they don't get adequate rest due to non-work factor such as crying babies and noisy neighbours, but I think that the FRMS is being used to relax duty restrictions while placing more of the burden on the pilot to self assess. This is a backward safety step.