One of the real-world availability issues that is almost unique to the F-35 is the planned reliance on synthetic training, with remarkably sparse actual in-air training. Whilst this provides cost benefits and the ability to train in a more classified environment it does present combat availability challenges, some of which are reflected in the comments above.
Clearly on operations a squadron needs a squadron's worth of combat ready aircraft with the honed maintenance, weapon loading, logistics and the support teams behind them. Back in the Cold War sortie generation rates were drawn from specific exercises and fleet experience. Quite simply, we trained as we fought when it came to maintenance, weapons, logistics and support.
With the aircrew element tripping the light fantastic in the networked simulator just how do you gain realistic experience and data for everything else?
The RAF/RN are looking to operate just a handful of real sorties in any given week. If a real tail number has not been utilised for a couple of weeks how do you measure its serviceability - do you simply count the days not used as 'serviceable' if it was ok on the last flight? If you crew-in to that jet and it fails and you switch to another aircraft just how do you record the previous history - was it really serviceable for the last couple of weeks, but not tested, or was it 'serviceable' on the basis that nobody had actually tried. If a squadron's worth of support personnel can be utilised to launch a pair of F-35s can you really turn 4/4/2 for 10 days+ with your sister squadrons doing the same?
Even for typical FJ ops the switch from peacetime requirements and manning to actual surge ops can be dramatic. In peacetime trades such as armourers have little to do and are often manned below a rather parsimonious set level. Switch to a large-scale shooting war and we run these guys to the ground. Same thing will happen in other pinch-point trades and large-scale ops cannot rob personnel from other squadrons as they will share the same commitment and pressures. Not every war is timetabled months in advance with just 1 of X squadrons deployed on a set rotation.
The long-term effect on maintainers with 'little to do' has a deleterious effect on currency, competency and any meaningful level of on-the-job-training. I can think of one current ME type that was deployed on ops for years without ever needing an engine removal since first manufacture. When the inevitable day came the engine guys had zero experience and zero confidence in undertaking the task. Thankfully RR and a civilian company were brought in to help with the engine change in the UK. With the F-35 the lack of flying hours will reduce the experience of all those who will be leaned on when synthetic is switched for real operations. Equally live ops may be the first time we stress components over a short timescale and there will be discoveries and new arisings that have just not been anticipated.
A reliance on synthetic training is a gamble and I argued long and hard that a platform had to be tested and mature at all levels before parking most of the fleet and heading for the (aircrew) simulators and hoping for the best. I lost the debate.