I understand what their attempt is but the problem is with the choice of the model i.e. F-35. If they had simply used a stealth design "model X" and apply whatever test based on with and without canards, then the conclusion is a much simpler conversation. Once you overlay a specific design onto the conversation, all the specifics then matters especially when the aim is to extend that conclusion unto the J-20 airframe 9by inference). The most fundamental flaw in logic is with the starting point, i.e. the F-35 and J-20 are of different designs. .Well, in certain way, I think you misunderstood about what they did and what they trying to achieve.
Let‘s look at the scenario here.
Firstly, I agree the general purpose of this paper may affiliate rcs impact over the j-20 canards, cause f-35 with canards just mean less because there is no such thing. And they might did the experiment on J-20 in the first place.
After they did the experiment and testing, they found, in general, canards with certain improvements could match the rcs performance of a conventional layout. But due to whatever reason, they realize the publish the original data on j-20 is sensitive or they have to through a lot extra efforts in order to do so.
So instead, they build a model for a fictional f-35, and redid the experiments and tests on that, and write a paper about it. Since this particular aircraft is not Chinese, it’s not real, so no secret involved, no clearance needed.
You see, the idea here is to publish the finding and method without specifying the data they use, as long as the conclusion is the same, they could live with the fact, that the real j-20 results may have slight differences, but hi, I don’t think they will disclose those data any time soon, in fact if you ask them, they may as well deny the connection between this paper to j-20.
Overall, the title of this paper is called “Study of canards impacts on rcs”, our guess is based on their background, but as an academic, when publishing a paper, their no.1 job is to get the paper published with the correct conclusion, the rest are all interchangeable.