I disagree. You've described a very good way of actually conducting science, which would probably be more reliable than the way we've been doing it. You italicised "disprove" in your earlier post because you know that when people do science, they don't try and disprove their own view. At best, they come up with experiments designed to disprove someone else's view. In reality scientists are searching for data that can be comfortably explained in terms of their view but which cannot be so comfortably explained in terms of competing views.
But the thing about science is, that even though people are doing it wrong, it still works. Science is somehow robust even to such fundamental mistakes, that it ceases to be science.
Then again, we might have spent the last century or so building a completely false model of the universe because our technique is wrong, because we haven't actually been conducting science according to the sort of definition you give. It's worth considering, but I doubt it's true.