The article touches on that, but I haven't seen any mention of testing revealing a large number of dioxins. That's why I hedged with "as far as what has been demonstrated".
I can totally see the scenario where the people signing off on the burn look at the ideal combustion products (only CO2, H2O, and a little HCl), while a much more imperfect burn creates a bunch of other crap, that becomes significant due to the quantities involved.
FWIW do you know which dioxins are possible? The only specifics I've seen seemed to be people fearmongering about things like 2,3,7,8-TCDD, which seemed implausible to form from vinyl chloride, at least as far as my rudimentary knowledge of ochem would suggest.
From a brief look at Pubmed it looks like the top articles investigating dioxin formation look at total dioxins formed, not the specific molecules. This one states that PCDFs form in greater amounts than PCDDs: https://pubmed.ncbi.nlm.nih.gov/17432330/
That's a great article for its breaking out the various chemicals and focusing on each one. In line with what I've been thinking, it's talking about dioxins as a result of burning PVC (which was in some bulk hopper cars, according to a manifest I saw), which was burning as a result of the derailment (as contrasted with the liquid vinyl chloride, for which the burning was seemingly a deliberate economic decision).
I agree that there should be more testing, and the fact that testing for expected combustion products wasn't start immediately to inform decisions about evacuation is criminal.
Dioxins are a result of low temperature burning of chlorinated hydrocarbons.