Earlier quoted context omitted.
I think the argument goes that if the civilization was sufficiently old and short-lived, no we probably wouldn't find any fossilized evidence. For example we've only found about one dinosaur fossil per 10,000 years of dinosaur history. The dinosaurs were around for a long time, from 65-250 million years ago. Humans have only been industrialized for around 200 years. If you imagine a dinosaur civilization that's indus…
In order for a civilization to be undetectable, you'd essentially bound terms in a terrestrial version of the Drake Equation [0] You'd need a civilization that did not produce long-lived technological signatures (e.g. glass panes), that did not have a large number of individuals (i.e. produce remains), that did not substantially alter their environment (i.e. leave geographic markers), AND that did not consume easily…
This is the topic of the paper, I think the authors settled on climate markers and radiation from elements like plutonium-239 as among the very few pieces of evidence that might still be around on that time scale. They also observe that there has likely been enough exploitable energy in the form of oil, natural gas etc. to support an industrial civilization since about 250 million years ago.