Dogma, what is there to investigate, and to what purpose?
The debate would be on a point of judgement; a continued landing in the same circumstances may have involved higher risk, so the appropriate ‘bounced recovery’ and GA was the better option. Yet a small change in the wind gust or a few seconds of flight path would present a completely different situation. Is it possible to predict that; I doubt it. (Hence you cannot simulate the appropriate level of fidelity G-CPTN)
How can you improve the outcome of a correctly executed procedure; can you better manage the risk, etc. Celebrate the success, reinforce the choice of a GA in similar situations.
If you wish to judge the choice in commencing the approach then you must consider the view of the Captain, the information available, at that time, skill, qualification and the many other variables. The human is usually very good at this type of judgement, superior to automation; and yet the same human is particularly bad in judging events after the fact, generating wild conclusions based on hindsight and constructing ‘error’, whereas the behaviour was as required for the situation.
Armchair views may be the greater risk to the safety in industry because they may covertly influence crews towards courses of action to mitigate media reaction vice managing the risks in flying.
You cannot conclude a ridiculous view of an upset, or inappropriate use of control; you were not there, did not have the same perception as the crew, or experience in judging situations.
You must start with the view that the crew choice and actions were as required; the crew were doing their best in the conditons, which from the video appeared to be by the book. Well done that crew.