If you examine their hype cycle charts over about a decade, you'll notice that technologies join or leave the chart randomly, and very few actually move in a linear fashion along the hype cycle: the charts offer no real predictive power.
I think the "hype cycle" narrative only matches a small fraction of tech innovations. Sometimes tech is adopted in a fairly smooth sigmoid. Sometimes it dies suddenly pre-plateau because it was actually vaporware or a substitute became more competitive. Sometimes there's a single giant hype cycle (dot-com?). Sometimes there are several repeating cycles (looking at you today, cryptocurrencies).