Can theories about the past be scientific even though the past cannot be observed?

The short answer

Yes, because the way things were in the past can have observable consequences.

The longer answer

In general, propositions about x can qualify as scientific even if x itself cannot be observed in the here and now, just as long as the propositions in question have observable consequences in there here and now. Were this not the case then, for example, the hypothesis that atoms exist would have been unscientific before the development of electron microscopes, since atoms are too small to be observed individually by less sophisticated means; the hypothesis was, of course, scientific long before then because the existence of atoms had observable consequences like some of the details of Brownian motion.

Since propositions about the past can have observable consequences in the here and now, they too can count as scientific if formulated with adequate care. Consider Big Bang theory, for instance. Big Bang theory makes a claim about the state of the universe some 13.8 billion years in the past. Although no one was around back then to observe what was happening, the occurrence of the Big Bang had consequences that we can observe today, such as the expansion of the universe, the abundance of light elements, and the existence of cosmic microwave background radiation. If scientists had failed to find any of these predicted consequences, then Big Bang theory would have to have been scrapped or modified. Fortunately for the theory, its predictions were borne out.

References

Ruse M. 1982. Darwinism Defended: A Guide to the Evolution Controversies. London: Addison-Wesley.