The entire purpose of VAERS is for researchers to check the data in there against normal rates so any outliers (i.e. potential problems) can be identified and investigated. VAERS is how we find unusual patterns that lead to unexpected side effects which would never otherwise be linked to a vaccine. With that information, warnings can be issued and people can be protected. So what you’re attempting to do is valid, but I can tell you that there are quite a number of people whose entire job is running those numbers constantly and they have more complete datasets. Not saying you shouldn’t do it; just pointing out that the methods are being continuously applied at scale and the results acted upon.
As for low quality reporting, you should check out the guy in the UK who filed a report that a vaccine turned his daughter into Wonder Woman. VAERS accepted the report with no question (as it always does). That wasn’t even someone in the US. At least the doctor who filed a VAERS report that a flu shot turned him into the Incredible Hulk (yes that’s real, and yes, VAERS accepted the report) was from the US.
That's 7 days / 365 * 0.008 * number of vaccinees / underreporting ratio. Assuming there are 4 times as many real events as reports, that's 3836 deaths in VAERS. So that's ballpark expected deaths assuming 100 million vaccinees.
For reactions other than deaths you need every shot, or 250,000 and typically day-of-vaccination or at most the day after. With 800,000 heart attacks annually / 330 million Americans that's 415 heart attacks with 250 million shots assuming 25% reporting into VAERS.
This is not difficult to understand nor controversial. Same numbers for stroke as heart attack, and I have not looked up the rest. That's what should be in VAERS, and any large excess of those numbers indicates unexpected adverse reactions.
For deaths, about the expected number of reports. For adverse reactions, more than expected, but I have to do more careful filtering to reduce the false positives and negatives.