To a large extent the problems with the F-35, and with most other current military procurement programs are due to the reversal of the experiment / development / production process.
In times of yore, a new product would be designed and built using the best possible available technology using information gleaned from development of earlier models and from pure research test beds. A new aircraft would be assembled using the best of the known off-the-shelf technologies, or those that were realistically well enough known to be a low risk. OK design optimisation would be required to get the best from a design, but in the main the lessons required to put a new model into service were learrnt by development on the previous model, or on pure research test beds. Result? new product development took relatively few (5-10) years and by the time the machine entered service there was a good chance that most of it was still "in date" technologically.
What do we have now? A situation where a company proposes a pipe-dream as being an attainable target, even though the technology does not exist. It gets a contract with a view to final production before the technology has even been proved. So what do you get? 20-year cycles before the aircraft (or ship or sub or tank or whatever) gets into volume service, by which time all the ancillary components are out of date.
Look at Harrier - when appraised of the basic Pegasus design, Camm had the basic design drafted out within weeks, using available airframe construction methods.
Compare that with F-35. Thats been built around an engine that didn't exist at he time and still has problems, with resulting issues which will need major airframe mods. Volume production is still ten years or more away, which means things such as the avionics fit, composite build materials, skin coatings will all be 15+ years out of date before there are enough aircraft around to get used. And thats before you assume a 25-20 year active life for these aircraft.
Someone needs to take a chainsaw to the whole military procurement process and separate the process flows of research and production into the two discrete streams they should be, rather than -as at present - carrying out basic research as a rectification process on an existing production design