But suppose we could not in the laboratory duplicate the conditions for chemical combustion or found that spectral anaylsis gave inconsistant results?
Then, of course, scientists would be making rank guesses about the formation of stars.
We have not been able to show in a lab how single-celled organisms can evolve into multi-celled ones. Much less with exclusively naturally occuring processes. Much less with consistancy.
That's fine. That's a great way of telling you what a star is doing right now, but that's not the same as seeing one form, is it? The animal equivalent would be to catch an animal and observe it in a zoo - that'll tell you all about what this animal is like, but it doesn't tell you where it came from, what caused this animal to come to be.
Or look at the meteor crater up above - nobody has observed impacts near that scale, so therefore, we don't know what caused it. Right? That's pretty much what the science teacher will have to say, isn't it?