There are 3 main issues here.Read the latest news about the JSF,"white elephant" claims.Still plagued with engine and software problems,costs rising.
Engine : Reliability metrics were measured. Out of the 3, 2 had the engine ahead of where it was supposed to and one where it was behind. The GAO did not take into account the design changes that were being developed, and/or certified to remedy that. They themselves claim that, that they will look at those changes and their impact later next year (report that is published in Feb of 2016) since they only book a change once it has been implemented and demonstrated. The program management claims that it is under control and the changes "work". It has no impact on IOC or weapons integration for any version. How the reliability of the engine has improved through these changes (or whether it has at all) will be known next year. The program took a decision to do concurrent development and in this case, if there are design changes (minor or major) found or required during development (this jet is still in the RDT&E phase) it is introduced back into the design. Thats how they improved a host of things such as mission availability which shot up beyond 65% late last year (for in development aircraft). Its an ongoing process and the process will continue till 2018 when the development phase finally comes to a close.
Software: 2b is on time minus the MADL 4 ship which will slip from July 2015 to October/November 2015. 3I (USAF) is basically 2b software ported to faster Integrated Computers and mission systems. That is flying as we speak and its expected on time since there is no software addition to 2b. 3F could be delayed by up to 6 months, but 6 months is exactly what they built into the "IOC WINDOW" (remember IOC between July (desired) - December (Threshold)" that is mentioned in every IOC claim. Even if they go over the 6 month window, the delay for 3F is likely to be measured in weeks above that so it hardly matters since it is nothing more than a software upgrade. They aren't changing wings, or rewiring the aircraft, or swapping sensors etc for 3F. The software will be flying in combat aircraft (all squadrons) starting 2017, they will get clearance to use it in an operational context only once the testing is complete and it is certified which would be towards the end of 2018.
COST: Cost has been coming down every block. Contract details are shared and you can look up budget documents each and every year when a new block is ordered. It was over $200 Million URF when the first jet rolled out, and is at $108 Million for the CTOL variant (A). The next order that will be announced towards this year's end will be for 150+ F35's, expect the cost to be close to $100 Million (net i.e. URF for engine+airframe and components) if not below that.
Risk remains in software (most significant), production ramp rates, and weapons certification schedule..that risk needs to be managed but its no where close to where it was say 2-3 years ago. If there is a delay in software it will really have insignificant impact since its likely to be for a few weeks or a few months at the most (above the threshold) and in the bigger picture that is hardly a big deal especially since its just for software and it would have zero impact on the production rate. You may essentially need to have hundreds of jets run and fight with 2b for a few months more. Thats about the size of the impact that may have.
Both the JPO and the GAO agree that unless the software development work becomes more efficient or speeds up they are likely to see a 4-6 month delay for the final 3F version of the software build. However, do note that at the program baseline they built wiggle room for the IOC deadlines because they knew that predicting exact dates is always something that could be delayed. The JSF isn't seeking a basic air-to air capability at IOC like the F-16 or Eurocanards, it is seeking full blown multi-role capability, with full sensor fusion, electronic attack/warfare and LPI data link sharing (something that the Eurocanarlds or the F-16's don't even have NOW) from IOC and then adding to that capability at FOC before embarking on the Follow-On-Development phase beyond that (in reality follow on development phase overlaps the RDT&E phase as block 4 software work and engine_enhancement work as already begun in support block 4 capability that is slated for the early 2020's).
BTW, Follow on development has already started for future capability and Bock 4 contracts were awarded in the Financial year 2015 budget, and requested in the FY16 budget. New weapons are coming to light such as the 150-200nm Ramjet powered ARM that is to be an extension of the AARGM that is in production. That missile received funding starting FY2015, and is expected to begin its operational testing by 2019 or 2020. Its a relatively simple upgrade since the sensor, and internals have been developed already and the VFDR was already tested on the HARM a few years ago (fired from a QF4). Additionally, a New Cyber/EW stealthy pod is being developed for customers (possibly the Marines here since they need escort and SO jamming because they aren't buying growlers to replace prowlers) if required to support the rest of the fleet.