WordPress caching is useful. It is not the whole performance model.
A cache plugin can make a stored response cheaper to deliver. That is valuable. But it does not answer the earlier architectural question: why did WordPress need to execute so much PHP, load so many plugins, and build so much context before a response could exist at all?
The cached answer still has an origin story
Page caching shifts repeated delivery away from expensive dynamic generation. But every cached document still had to be created once, invalidated later, regenerated under certain conditions, bypassed for specific requests, or missed when the request does not fit the cached variant.
That is why caching should not be confused with workload prevention. Caching can reduce how often WordPress repeats work. Workload prevention asks whether parts of that work should have been started for this request in the first place.
The missing layer is execution scope
Most optimization advice focuses on output: files, scripts, images, render paths, Core Web Vitals, and cached delivery. Those layers matter, but they begin after the request has already awakened WordPress and its plugin stack.
Rush - Powered by LiteCache approaches the same problem from an earlier point in the chain. Its principle is simple: do not execute what the current request does not need. That does not replace caching. It completes the model before caching begins.