The ongoing studies being done by the DOE are showing that embrittlement should not be a hurdle. The problem is that behavior curves had not been developed for 80 years of aging for the materials of concern. Ongoing experience plus R&D allows for extending those curves through proven methods. Still more work to do. Of course, if needed, simple annealing can take care of embrittlement. (simple in concept, quite a bit more challenging & costly to implement in the plant.). Cracking in welds can be easily seen via existing examination techniques, and repair methods already exist and have been used.
Concrete is a bigger challenge. More work needs to be done on irradiated concrete aging. That work is in progress. By the time the first plants hit 60 years, that data will be solid and reliable for understanding longer term behavior. That work is important.
Those are generic aging issues. Each individual plant could have specific issues of its own, like aged cabling, so the process of renewal must be thorough. Fortunately for the industry, great margins were designed into the plants to begin with, and many of the critical components can be replaced in part or in whole.
What will likely play out is the some plants will retire at the end of their extended service period, just like any asset. Some of those might shut down a bit early if they have equipment problems that are too costly to repair for ongoing safe operation. Those that have the best performance will be likely candidates for another license renewal, which is typically 20 years. The utility industry is wary of becoming too dependent on gas. Extension of safe operating life of nuclear plants and/or new nuclear are the only viable options assuming coal is dying, or unless storage for wind power can be made economical.