"*"Limited data" - Actually massive numbers vaccinated with health authorities scrutinising the pharmacovigilance systems."
That data is still necessarily limited. This is a site that carries mostly news about technology, so let's talk about something clear and specific from our own industry. When we manufacture computers or other electronics, we often perform accelerated aging tests on samples prior to market introduction. This allows us to gain some imperfect information about how long we can expect them to last in the field and what the distribution of failures will look like over time. This information can be used to find and address defects, improve the design, and establish economically viable warranty terms. Because we need or want, for economic reasons, to get the product into the market long before its expected or required lifetime has passed, we necessarily take shortcuts in testing. Some tests address things like design defects (always present) and manufacturing defects that manifest themselves on every device more or less immediately. Others address statistical problems that affect small quantities within large lots -- think mechanical tolerances, marginal soldering, slight footprint mismatches, etc. But to get to market when we need to, we also employ accelerated aging tests that attempt to mimic some or all of the things that happen slowly in the field. These tests are necessarily imperfect but they do have some predictive value: they have been shown to identify defects that other tests cannot, and to have some correlation with actual field performance over time.
There are no accelerated aging tests for human beings, and if there were, it would be unethical to administer them. The only way to know the long-term effects of a medical procedure -- including a vaccine or any other drug, surgery, etc. -- is to wait and watch. If you want to know the 5-year rate of side effects from a specific vaccine, you must administer that vaccine to some large, diverse population and a placebo to some other large, diverse population, then wait 5 years and perform occasional examinations of both populations. There is no shortcut. One cannot simply spray humans with salt water for a week and draw useful conclusions about side effects over 5 years (or some other arbitrarily long period of time). That means if you want confidence about what side effects (if any) occur after 5 years before introducing your medical treatment into the market, you have to wait at least 5 years after you first develop it and begin large-scale clinical trials. Even the very limited short-term trials most regulators require for drugs and vaccines are extremely expensive; the kind of long-term study I'm describing is mind-bogglingly so. And of course there is demand for treatments *now*, both from patients and sometimes from government officials; many people don't want to wait 5 or 10 or 25 years to understand the long-term side effects, especially if the treatment addresses a serious illness they already have, or if they've been whipped into a hysterical frenzy of terror by the mass media, politicians, and corporate lords looking to score policy gains.
The end result is that while the data on the short-term side effects of most Sars-CoV2 vaccines available in Western countries is fairly extensive and does in fact show that harmful side effects exist but are very rare, there is no data on medium- to long-term side effects (if any). This is not because their manufacturers are evil or government officials in cahoots with them (either may or may not be true, they just aren't relevant), it's because the necessary amount of time has not passed. Unfortunately, there is also a very strong taboo around funding or performing long-term studies of vaccine safety. Many people, rightly or not, have concluded that this is because some vaccines are much less safe than the establishment wants people to believe, or because those establishment entities simply don't want to know. As a result, although it's conceivably possible to perform these long-term studies, it's rarely done, especially since the Wakefield paper. Therefore the true long-term safety of most vaccines is simply unknown; in some cases because it's unknowable, in others because it's not been studied adequately. One of these problems can be fixed; the other cannot.
In the case of all Sars-CoV2 vaccines, the long-term side effects, if any, are unknown and unknowable. I would say that qualifies as a "limitation" of the available data, wouldn't you? It's fine if you believe either that Sars-CoV2 is so dangerous to you that this limitation isn't important, or if you choose to believe that all vaccines, or at least the Sars-CoV2 vaccine you choose to receive, are universally safe in the long term. There's no evidence for that, but there's no evidence for the existence of gods either and we don't punish people for being religious. It's also totally reasonable for someone whose actuarial life expectancy is a few years or less to decide that long-term thinking simply isn't applicable. I'm not going to tell you otherwise because I don't have any evidence to the contrary -- no one does nor can -- and in any case decisions about your medical treatment are yours to make and yours alone. But I do ask that you accept that there are in fact limitations to the available data that might lead reasonable people to choose not to receive one, or any, of the available Sars-CoV2 vaccines.
Actually, I don't ask; I insist. Accepting these facts does not make you an anti-vaxxer; it makes you an open-minded human being who treats others with respect and decency and acknowledges the omnipresent possibility of being wrong. The alternative -- whether you're an anti-vaxxer or a corporatist-statist propagandist -- is to be an obnoxious bully. That's what Meta are doing in this example, and what far too many of you are doing here. I don't like bullies, and I stand up to them.