I've seen listings on VAERS where it was "Death due to car accident" and things like that
You'll find similar cases if you delve into the CDC and equivalents' reporting on Covid deaths. "Death for any reason in X time period after confirmation of Covid" is the norm. They specifically state that they count all deaths just so it's clear that oversampling can be a possibility, but it's always been preferable to undersampling. It's just that the extreme politicization and polarization that Covid has brought with it makes any discrepancy or vagueness in data look suspect and possibly malicious in the eyes of an agitated public.
As others have said, it's just good bookkeeping on the end of the medical establishment. I'll add that the parameters under which research and statistical are done are of vital importance and in practice are very often not as pure as people like to imagine.
@Secret Asshole is absolutely right in saying that the validity of any given research paper is in the minutest details. A good study and a deeply flawed can look virtually identical, with the one crucial difference being in how they conducted or phrased a certain element of the equation, which will fly over the heads of most laymen.
In the case of Covid/vaccine statistics, the simple fact is that the bodies tasked with compiling them are usually receiving mountains of second-to-third hand or impure information, and are expected to sift through and publish it on a day-to-day basis because government policy lives or dies on it. So, combine the aforementioned standard operating procedure of academic caution with the above normal level of pressure, and you get a methodology which purposefully casts its net wide for reasons of safety and expedience. The Internet might mock or rage at the odd vaxxie/coofer who got hit by a truck and made it into the list, but the thing is that for every one of those cases, there are 10.000 more which are considerably more valid that would've been overlooked had the criteria been stringent.
It's similar to how some vaccists will try to discredit the notion of vaccine side-effects by pointing out that most cases are self-reported, while in the same breath hyping the horror of Long Covid when data on that condition is collected in the exact same way (or vice versa). 90+% of complaints in either case are going to be mild and unverifiable symptoms like 'malaise' or 'fatigue'. It's not practical or germaine for the people receiving the intel to go back down the line and order biopsies for everyone or some shit. Better to focus those efforts on the complaints which are both serious and practical to investigate. It's likely that it would've taken considerably longer to discover certain symptoms of Covid, or side-effects of the vaccines, if medical authorities had been more discriminate in the type of information they chose to accept.
People on either side trying to dunk on the other by nitpicking universal methodology are exceptionally dumb, because it's the one constant to all of the credible sources they all use. The only thing such an argument demonstrates is the bias of the one who uses it.
