In flying, you file flight plans. Air traffic controllers make sure they are followed and orderly. All passengers are highly screened before entering secure areas and employees/food is similarly screened before making it to a terminal. Some flight officers need thousands of hours of experience. When planes are repaired the procedures are exhaustively documented, need to be done by qualified staff, and the replacement parts need to be traceable.
In the case of neurological side effects of COVID , it's largely an unknown and it can be reduced immensely by wearing a $0.10 piece of cloth, modified social interactions, and contact tracing (see most developed nations outside of the US and Sweden). COVID risk mitigation is much easier.
To quantify it a little: the article indicates a roughly 4x increase in a particular severe neurological condition (ADEM) over background incidence rate, and serology indicates < 5% of population exposure in most places. The 4x rate is on adult ADEM cases, and is very likely just [a subset of] cases that pass a threshold on a distribution. It's reasonable to spitball a 50-100x increased risk of neurological side effects. It's probably much higher because of the well documented clotting disregulation and the spectrum of mild to lethal strokes, and the structural damage strokes cause beyond that.
Ultimately, since the scientists in question are experts, I trust them instead of thinking I know better. I was wrong about mask efficacy early in the year, and that was humbling. Most of the scientists have studied this stuff for decades and the best strategy is to just listen.