If it's not a bit of partially-digested cheese, it might have to do with the interface between relativity and QED.
But while it's not unthinkable that relativity requires revision under some circumstances, it's one of the most thoroughly and repeatedly proven tenets of modern physics. It's going to require something much more definitive and reproducible to overturn that mountain. I wouldn't expect a positive answer to this inside ten years, but a negative answer could turn up any time since it depends on someone finds the bug in the gears.
That's one of the big steps which is necessary. It still leaves the questions of (a) funding, (b) how do you reproduce it and how long does that take, and (c) how do you explain it.
Many problems in science are considered to be "twenty year problems," by which we mean that they'll probably be solved eventually, but that it's unlikely to be any time soon. Many such problems go on for much longer without a solution despite significant progress (see also: fusion power plants). This isn't quite on that scope - but it could evolve into one if it actually ends up with a result which can be reproduced on command.
But while it's not unthinkable that relativity requires revision under some circumstances, it's one of the most thoroughly and repeatedly proven tenets of modern physics. It's going to require something much more definitive and reproducible to overturn that mountain. I wouldn't expect a positive answer to this inside ten years, but a negative answer could turn up any time since it depends on someone finds the bug in the gears.