The simplest split with potential to start addressing concerns in this particular lawsuit is to break the index that results from google's web crawling out into a utility.

With competing "engines" (defined as a ranking algorithm and frontend to query said algorithm) building from the same, high-quality index competition in the search space could get much better.

Engines such as DuckDuckGo relying on Bing for the majority of their index is a decent example of how this might work.