You're getting downvoted because you're fearmongering based on nothing more than "Everything is connected!"
That's not an argument, that's handwaving things away in a panic.
While it is not impossible for things to go wrong, the rational belief is that it will — to a higher standard than most people would consider “certain” — be both safe and protective.
The alternative scenario here is having those same viruses and RNA inhaled in an active form with lungs helping to transfer them into the blood stream. What are you actually advocating for/against here?
Such a thing cannot be determined easily and effects can start showing after many years of the body not getting the flu.
You belive that it's easy to see how the body works: modern science can tell that only in isolation, but the body as a whole is not something we can predict the longterm effects for. Why do you think vaccines need 10-15 years of testing before getting approved (except the covid one, which tells you a lot about how much they know about the side effects....)
Because most diseases are much slower to propagate than covid. So you need to wait very long to have enough people that got it in your control group to have significant statistical results.
I'm no doctor but I feel like modern medicine can make some pretty darn good predictions.
Besides, your arguments are not sound. If your premise of "the body is too complex to predict" is true, then you can't immediately follow that with a prediction of a long term effect (e.g. eliminating a "harmless" disease will have ill effects.)
I can just as easily make something up like, "the immune system can only respond to a finite amount of disease, therefore getting vaccinated leaves more capacity for future infections." (which is not true)
And no modern medicine cannot predict anything on a global level of the body: if it could then we could've simulate billion of virtual humans using supercomputers and go for real human trials only after that. But such simulations would be innacurate just like how weather prediction is innacurate for more than a few days in the future: the problem is that the body (just like the weather) is a complex system where the butterfly effects becaomes a real thing. So it's not a question of "models" or "compute power": even the slightest devition in the model will break all predictions results in a cascading manner over time.
Bureaucratic delay and running trials in series, mostly.