The "Accelerate" book claims it is not just correlation but even causation: Higher deployment frequency causes better company performance.

Given that its trivially easy to produce unbounded amounts of useless PRs this cannot be a causation.