Lagging indicators are not useless; they are confirmatory. They tell you a regime change was real rather than noise, which is exactly what a long-horizon allocator needs and exactly what a trader cannot wait for.
The danger is policy made on lagging data. Inflation and unemployment both turn late, so a central bank that waits for them to confirm has, given policy-lag, already overdone it in both directions. This is the core argument for acting on forecasts instead.
Example: the recession peak is later dated to month zero. Unemployment does not rise half a point until month seven and core inflation does not decelerate meaningfully until month fourteen, by which time the cycle is nearly over.
Related: leading-indicator, unemployment-rate, policy-lag, core-pce, nber-dating