Maybe guitar makers are not innovating? Other stringed instruments have basically remained the same for centuries, but electric guitars were born out of modifications to standard guitars. Hendrix was a guitar hacker. Most of the electric guitars on the market today are no different than those sold 30 years ago. There are a few new ideas (midi accessories, built in effects, the Moog guitar, more than 6 strings, Steinberger's, cryogenically treated strings, better tremolo's, guitar synths) but few people are into hacking guitars any more. Your mom and dad may play electric guitar, it isn't radical or rebellious. And few are trying to invent new styles of playing either. So the sound and use of guitars has remained fairly static for a long time (except for Fred Frith!). Perhaps this explains the recent interest in Eurorack synthesizers? A return to creating fundamentally new sounds by tinkering with the basic elements of sound.
It's kind of like the smartphone market, the manufacturers have tried all the easy improvements (bigger screens, faster CPU's, better GPU's, thin bodies) but haven't tried the more difficult things (unusual sensors, regional peer-2-peer communication, serious FDA approved medical sensing/apps, light based communication, UWB links into IoT networks, etc.) Though AI chips may result in some significant new ideas. Maybe deep learning AI chips for guitars would rev the market? You teach your guitar some notes/chords and then it can generate different arrangements/timings/bends for you that actually sound interesting? Maybe not since deep learning can only mimic what it learns.