It is possible that we are a brain in a jar being fed sensory data.
This idea is part of Descartes' Cogito Ergo Sum (I Think, Therefore I Am). We can't prove that the sensory data we perceive is an accurate reflection of an external reality. Descartes argued that the only thing we can truly be sure of is our own existence.
Even that might be going too far. We may have just popped into existence (or been created) with our memories of a continuous history simply being part of the now and not a true representation of continuous thought.
Even at face value, optical illusions show there is some flexibility in how we interpret sensory data. And everyone who has misplaced keys knows that memory can be fickle.
This just means that we can't be (or shouldn't be) completely and utterly certain about anything - whether an experience or a thought.
For the most part we accept this lack of absolute knowledge and make do with our best guess at what seems most plausible.
Science has a long history of our best guesses at the time. Some of these guesses have been refined over time, others have been rejected in favour of new ideas.
None of the ideas can be proven true beyond all possible doubt.
Likewise, nothing can be proven wrong beyond all possible doubt. While we may have the idea of 'impossible', we can't prove any given idea, concept, activity or model to be definitively impossible.
While mathematics takes the principle that something that is both true and false is inconsistent (inconsistent is the mathematics version of impossible); this isn't a proven principle (it is a plausible principle - but not proven beyond all possible question).
This doesn't mean that all ideas are equivalent.
Science has evolved a number of principles that help us evaluate ideas.
1. Can it be communicated? However magnificent the thought appears in your head, nobody else will be able to consider your idea unless you are able to communicate it in some fashion.
2. Is it testable? Does it make predictions? We can say all sorts of things about what exists outside our universe - but if we can never measure anything outside our universe, those ideas can't be evaluated.
3. Are the results repeatable? If nobody else can emulate your results then many people will be skeptical that you achieved the results you claim.
4. Occam's Razor. Given multiple ideas that fulfill 1-3, we prefer the one that is sufficient to explain the observations without being more complex than necessary.
None of these points is enough to definitively prove or disprove something. The man with your brain in a jar could be messing with you and making all those solipsistically illusory disbelieve you for the chuckle. Likewise, the brain-jar guy could be making it seem like those illusory solipsistic people are agreeing with your observations.
1-3 are required in order for there to be a productive discussion.
As much fun as it can be to speculate on what can be done using the impossible - it isn't possible to conclude anything one way or the other.
Proposing ideas that cannot be communicated isn't wrong in itself - it just isn't productive.