The founder effect differs between small populations filtered from a large population vs a smaller parent population.

I want to point out that the silurian hypothesis is specifically about civilizations removed from us by geologic time. Advanced human precursor civilizations are much more effectively ruled out by the current evidence. Basically, a hypothetical precursor society either has to look very different from any we know about and materialize out of nowhere like alien colonists, or they didn't exist.

For example, we know they weren't using large amounts of fire because nothing in the global charcoal records points to that. We know they weren't using plants that produce pollen because we don't find phytoliths with domestication indicators. We know they weren't using surface metals, because there was lots of that still around when the classically ancient civilizations started. We know they didn't contribute significantly to the gene pool. We have no material culture from them. We don't have a location on Earth they would have lived in. Etc.