You can't ever test anything in the past, so by your logic nothing in the past can be considered scientific. By your logic, plate tectonics can't explain the past movement of continents.
Which is, of course, wrong.
We can absolutely test aspects of a theory, verify its
general validity, see that it best correlates with past evidence, and therefore infer past causes. Nothing of that is unscientific -- to the contrary, that's precisely how science works.
If all we had was observations of modern day plate tectonics it would be foolish to back-propagate the same motions billions of years in reverse and claim ones methods were "scientific".
Luckily, we have things like fossil records to justify such conclusions. They provide the second point with which we can "fit a curve" when exploring history.
With this "origin of life" stuff, we don't have that second point of confirmation. We think up a bizarre process, execute it, and see that it produced some molecules that we see in modern day life. We then have the gall to say that we've discovered "the origins of life"? Yeah right... https://news.ycombinator.com/item?id=27974369
If I may use a programming analogy, the process described is akin to NPC agents in a browser based MMORPG Rowhammering a bunch of random bits into some exposed ArrayBuffer until a Counter div appears on the page, then claiming they have some deeper understanding of the origins of life in their universe.
Do they have a deeper understanding? Perhaps. Have they come anywhere near the "origin of life"? Nope.
