As an individual example: in 1998, Intel's top-of-the-line CPU was the Pentium II with about 8 million transistors. In 2008, they released the Core 2 Extreme QX9770, with about 820 million transistors; this is a 100-fold increase in 10 years, which is almost exactly what Moore's law predicts.
Fast-forward another decade to 2018, and Intel's highest-end Xeon chips have something like 7 billion transistors, which is less than a 10-fold improvement in the same timeframe.
EDIT: that said, I think it's a real stretch to go from "the slowdown of Moore's law" to a sweeping claim about innovation in general being more difficult. There are a ridiculous number of different fields in which someone could start a business; relatively few of them are dependent on enormous, ever-increasing amounts of computational power.
I'd love to see a production cell phone that is already thick with a relatively huge battery, more rugged case, and a slightly better protected display. I'm willing to bet there's a market for this, just based on otterbox sales, and the sales of backpack battery covers.
Similarly a home media player like the roku, shield tv or fire stick are plenty small. That has nothing to do with being able to innovate and create newer or better experiences.