>> Take for example acetaminophen (aka paracetamol/tylenol): it is borderline ineffective, has been superseded by ibuprofen for almost all use cases, and the required dosage for its tiny effect is also very close to dangerous levels.
This is only really accurate in a small window of analysis. It's been shown that paracetamol + ibuprofen has synergistic effects for most pain-related issues and paracetamol is superior to ibuprofen for migraine headaches; furthermore, the obvious correct headache drug of choice is aspirin/paracetamol/caffeine, aka Excedrin, which is vastly superior to ibuprofen.
I won't argue that paracetamol has acute liver toxicity issues (the largest drug of overdose problems, though a lot of this is intentional rather than accidental), but ibuprofen is abused in a chronic fashion with buildup over time leading to renal failure and worse (pro/amateur athletes are particularly at risk of chronic abuse issues).
I won't argue that paracetamol is way more dangerous than people think, and there should be far more warning labels (which J&J fights against), but I would be genuinely shocked if it got its approval revoked; similarly, I'd expect it to be approved if it was around today as well - especially if we knew about the synergistic effects it has with aspirin and caffeine.