Except this paper is not a review. Here is a review: https://arxiv.org/pdf/1709.0492
Not good that cross sections calculated by perturbative QCD change so much from LO to NLO to NNLO and now with N3LO corrections. This does not give much credence to the underlying model. Furthermore, this only applies to high-energy physics, because perturbation theory and with it the concept of force-mediating particles fail in low-energy QCD, bound states, solitons, etc...
Perhaps we use the word review in different ways?
Your review has 23 references.
My review has 217 references.
Length, number of citations, vary similarly.
More specifically., your review is looking at one particular issue in the comparison of experimental data and one specific set of MC generators used to perform computations in lattice QCD for specific LHC 2017 analysis.
My review is looking at a general change in methodology (adding NLLx small-x resummation) that improves fit across a wide range of experimental datasets - comparing this with previous work.
I don't see (other than linguistics) any argument, except that your review is clearly not suitable as an overall review of the merits and demerits of 2017 work in this area. Maybe mine, because focussing on the NLLx resummation, is also not suitable, but it does give more insight.
If you argue that QCD calculations are so mired in complexity that it is difficult to be sure they are right - then I'd agree with you. This however is a criticism of the computation and not the theory. It does reflect badly on the theory to the extent that the uncertainty makes it less predictive in those areas where calculation is difficult (most). So I accept your concern that the calculations are not great, but not that the underlying model is therefore wrong. I'm also not saying the underlying model is the final word, or even particularly satisfactory. But, however messy, it has a lot of predictive power other than the computationally horrible determination of parton distributions in quark-heavy space.
One way to think about this is that SM provides a unified and extremely successful model for all known particle physics - with the exception of dark energy and matter that seem necessary from cosmological considerations but are not (currently) relevant to particle physics. In this model precise calculations work well except for QCD where the perturbative expansions are not nice for understandable computational reasons. That makes QCD less predictive (at the moment - unless/until we develop better computational techniques). When it becomes more predictive maybe it will prove more clearly wrong or right.
Physicists crave simplicity. I've never though that it is reasonable also to crave computational simplicity. Why should God have made the workings of the universe easily calculable? you might almost (in a theological context) take extreme computational complexity as God's way of giving us all free will - though this is not a theological site and many people would less subtly invoke HUP as that.