I would say one shortcoming of this paper here, which is that it does
not justify the use of Table 2 in Level-1 as a good and reasonable
profiling approximation, the method seems plausible, but more and
in-depth explanation may be needed.

Hongli Zhang