Science is a methodology. It makes no claims to producing any final truth. It does, however have the capacity to debunk fraud and error in claims about the way things work.
It does produce statements about the world that are more reliable than armchair philosophy, even if limited in scope.
Individual scientists can be armchair philosophers and say anything they want. they can be wrong in that same way and for the same reasons that non-scientists can be wrong.
But within the scope of observable phenomena, science produces better and more reliable statements than any previous method.
Science is a methodology.
And methodology is what now? (Note that definition you never responded to a week or so back)
"But within the scope of observable phenomena, science produces better and more reliable statements than any previous method."
And does non-observable phenomena exist?
Must something be credible to be reliable?
How can one rationally say that they can divorce the concept of truth from the concepts of either reliability or credibility?
Challenge: try saying something reliable about a readily observable phenomenon that must not at the same time be truthful to make it credible.
Next Challenge: try saying something false which debunks fraud.