Genuine experimental music would have to be a methodical investigation manipulating one variable with the end result remaining unknown until the experiment is complete. Experimentation is a process, not a result.
An album of experimental music would not be a bunch of tracks with odd noises or unusual time signatures - it would be one track played at a range of time signatures (some of which would have to sound terrible in order to find the boundaries). Which may hold some interest as an oddity but seems less compelling than the conclusion to the experiment, ie. the one that sounds best.
Save for the rare cases where the process of experimenting holds up as a result in it own right, surely the experimentation should happen behind the scenes: Musical Science pushing the parameters of sound to concoct a final result based on those findings.
Not a Genre but a process.