Well yeah, the illustration on some website (that may or may not actually be in her book) isn't scholarly evidence, that's true. But you're not giving her enough credit here, by a long shot!

1. She's been doing this since 2011, her TED talk a decade ago racked up 2M views, and this book landed a positive review from the curator of the Smithsonian's "hall of human origins"; she's not some rando. Here's her (somewhat outdated?) Google Scholar page: https://scholar.google.com/citations?user=QaDkX_UAAAAJ

2. The actual evidence here is supposedly "a unique database that holds more than 5,000 signs from almost 400 sites across Europe". It could still be misleading of course, but it's a lot more than just a website diagram.

3. You putting "language" in scare-quotes is completely unnecessary, as that's not what she's arguing at all. Rather, she's saying that these symbols should be treated as a milestone on the way to written language ("first indicators of our human ancestors capacity for symbolic meaning"), not full grammars in-and-of-themselves. Given that evolutionary linguistics is in a "pre-Gallilean" phase at best (to quote Chomsky), I'd say any well-cited contributions to the field should be welcomed! Maybe she's wrong, but in a way that leads us to what's right.

I came to the comments to be dubious as well, so I appreciate where you're coming from. But IMO "ridiculous" is going way too far...