Oh, that's a given. Your instinct here is right.
Has to do with society functioning as a genetic algorithm: look into a-life like the work of Danny Hillis. You don't make gains when one 'solution' dominates over all others, that's called a plateau. Currently we're at a very low plateau where meddling with finance instruments or collecting rents of various kinds is the only effective way to wealth, and where most innovation will just get you broke and out of the game, because the risk of failing is far too severe.
In history we've had far better conditions for innovation (for instance, the dawn of personal computers) when people were generally more well-off but the disparity between extremes (Gini) was far smaller. This wasn't true for all populations, but I mean the populations that supported technological innovation.
You have to have whole thriving populations of individuals with bits of the eventual answer, including bits that aren't in themselves self-sufficient, in order for crossover and inter-pollination to give you new heights of innovation. Part of that is challenging (disrupting) existing 'best answers', something that's already in the DNA of silicon valley. Right now the best answer is just to get wealthy by any means necessary, preferably by collecting rents and bringing nothing to the table (to work or develop new things would be redirecting energy that can be used to just collect rent and extort money out of the system: it's disadvantageous to make anything new or better when you can engineer the existing system to your benefit)
There is no way Gini inequality helps innovation at all. Innovation has to do with populations of actors that are all reasonably capable of addressing a problem, not seeking a single optimal individual who will do all your innovating for you. Our incentives are literally backwards.