Creating a new chip is a huge endeavour, they have to choose focus carefully. Current mobile CPUs are "fast enough" for everyday tasks (switching to another programming language rather than just optimizing the existing won't make that much difference, they are very unlikely to prioritize C++ since it's not widely used).
I think they are going to target smaller/thinner, more integrated devices with better battery life, along with continued innovation, like better integrated Soli and features enabled by precise motion tracking, at the same time making it so they can get away from Qualcomm eking out mainstream features. Absolute features is more of a checkmark features for enthusiast sites than generally meaningful, especially since there are so many factors like GPU, cores, interface speeds, etc.
It's really only some gamers that care about the highest performance. Can't find very specific stats, but according to what I found, 10% of gross screen time is mobile games, of which I'd guess the vast majority is low resource games like crosswords. It's still a huge industry, but Google might focus on Stadia for the more demanding market.