My overall point is that you cannot rely on a design document to tell you about the characteristics of an aircraft. Over its development process, aspects are changed and flying characteristics may disappoint.
The idea that since the J-20 is supposed to be maneuverable, it is maneuverable is intellectually unsound. The reports we've obtained suggest that the J-20 in actuality is not significantly superior than the J-11 and J-20 is the subsonic domain, but that it outperforms in the supersonic domain. But think about the implications: the F-22 and Su-57 are known for exceptional subsonic agility. If the J-20 is merely above average, it implies that in the subsonic flight domain it's not as agile as the F-22 and Su-57 in that region.
This is not necessarily a bad thing; remember, with future 180kn engines, a 16000 kg empty weight, and an estimated 28500 loaded weight, the J-20 has a fueled T/W of 1.29 and a 60% fuel T/W of 1.55 .
I think I've used this anecdote before, but consider this: a A6M2 in its ideal flying regime can possibly outturn an F-16. But one of the reasons the A6M2 isn't an air superiority fighter of choice is because its flight regime is easily supersede-able; i.e, the F-16 can go faster and stay faster than the A6M2, so even if we ignore air-to-air missiles the A6M2 is what it is, a World War II aircraft.
In the same way, the J-20, if the report of empty weight is credible, can force engagements at higher speeds than what its rivals are comfortable with.