How hidden would a J20 stay if it provides this kind of targeting which I guess would need to be updated multiple times during the approach of the missile ?
I doubt there is anyone who can definitively answer that question. It will depending on a whole host of key factors, such as how well the J20's AESA radar's LPI mode works; how well opfor's RWR and EW suits work; how securely the J20's datalink works etc.
How to use the J20 to minimise the likelihood of detection while maximising their impact on the battlefield will be something the first J20 pilots and the best minds the Chinese have available will be working extensively in the coming years to find out.
It took the US about a decade to be able to use the raptor to full effect, and I don't think it would be unreasonable to say that would be the benchmark amount of time the Chinese would also likely need to fully master the J20.
Unlike the poor F22, I would expect the J20 to get a comprehensive upgrade and improvement programme to support it throughout its life, so the PLAAF may well be chasing a moving target with mastery of the J20, as successive blocks get all the latest innovations and breakthroughs incorporated into them, giving the PLAAF pilots the welcome challenge of learning how best to use those new tools and features to full effect.
If they have not done so already, I would expect the PLAAF to pitch the J20 against all existing fighters, AWACS as well as ground and sea based radars to see under what circumstances would LPI emissions from the J20's radar, and the J20 itself becomes detectable to different radars.
Then it becomes a mini-arms race, with one side of the PLA trying to figure out how to reduce the likelihood and ranges at which the J20 becomes detectable; while the other side works to see how they can increase the likelihood and ranges at which they can detect LO opfor targets.
The Su35 purchase may well also have been done partly to help provide the J20 with as close to a foreign 5th gen as possible as a benchmark comparison.
So, after testing the J20's ASEA's LPI mode, it would be very useful to repeat the test with the Su35's radar in LPI mode to see what the differences are.
Similarly, you can have a J20 light up the Su35 with its AESA in LPI mode to see if and at what ranges the Su35's EW suite realises it his being scanned. Could reverse the roles and see how the J20's EW suite responds to LPI scans from the Su35 etc.
When they find differences and unexpected results, that's when the hard work will begin to find the root causes and to use tactics and/or design changes to mitigate them.
Now I hope it starts to make sense why it might take a decade for the PLAAF to achieve full mastery of the J20.