Intel's problems stem from the company being controlled by financials and not engineering. They focused on diversification because there was no more market share to gain with x86.
Delays in 10nm didn't matter, it wasn't like there was any real competition. What? You want to do X to fix the node? That will cost a lot, it's best to just refine what we have and focus on 7nm, that will be easier and cheaper.
Instead of trying to innovate within their own market, Intel tried to diversify and kept buying up start-ups, practically all of which have been failures from the outside. Everything was fine and dandy so long as they made money and the share price kept rising.
Now Intel needs to rebuild its culture, it needs engineering leadership and enough paranoia to seriously be working to create the best processors and innovate beyond adding x performance to their CPUs per year.