Well, it should go without saying, but there is a vast difference between shouldn’t and couldn’t.
You shouldn’t load something beyond its stated max load capacity, but all good designers build in at least double that load tolerance in their designs. It’s very basic contingency planning.
If Yang Wei truly thought agility doesn’t matter, he would have came up with a radically different design. The big costs to make the J20 a good dogfighters have already been spent in terms of the very design of the airframe itself. Compared to that, the extra weight from a gun is peanuts.
of course J-20 could dogfight well, but that does not make it a dogfighter, especially J-20 is definately not intend to dogfight as much as f-22 and su-57,
make one able to do something, or even good at something, doesn't mean that certain something was the one all intend to do, so your logic doesn't make sense, J-20 can dogfight well is a sufficient condition to be a good dogfighter, but non-necessary condition to be a dogfighter
also, most importantly, you don't understand the value of agility in BVR right?...
let me remind you, why J-20 pilots keep saying, J-20 is optimize for its supersoinic performance, in subsoinc, J-20 performs just as well as J-11C and J-16, but it completely dominate the supersoinc. and what's the use for supersonic agility? BVR
The current set up without a gun is good enough, but why settle for good enough when you can get ‘perfect’ at almost no cost once you have the (weight) budget for it with the WS15?
this is not a good reason to add something that could be completely useless, if it change to WS-15, under the same speed, J-20 could have higher top speed, higher cruiser speed, better payload, better agility and etc. these could make j-20 fly faster and even better at BVR, which are much more valuable then adding a gun to j-20 and make it dogfight