All those downsides apply to wood and aluminum to varying extents. It's a typical case of engineering trade-offs.
Wood chips and cracks (well, splits but it's basically the same thing) dents in any metal structure cause stress rises that do basically the same thing for strength as chips and cracks in fiberglass.
Wood and metal assemblies will all "fail suddenly" when overloaded because a huge part of their strength comes from their shape, they deform from overloading and that reduces strength (positive feedback loop).
Regarding heat, a quick google search turned up [1] which is good to just shy of 400f, about same temp at which the strength of 6061 falls off a cliff. I also found this[2] which to me indicates that the underlying chemical science is at a point where a fiberglass should be able to replicate the heat resistance of aluminum.
https://www.fibreglast.com/product/high-temp-epoxy-resin-300...
https://chemistry.stackexchange.com/questions/22198/are-ther...