Google's John Mueller responded to questions on Reddit about the figcaption tag for figurines. The NY Times uses it for their images and SEOs want to know how Google treats it. Google does read the content within figcaptions but they don't treat it as alt text, they treat it as a caption.
John added that image search does pick up on both captions and alt-text, and since figcaption is treated as caption, image search can pick it up.
Here is what John wrote:
I think what's happening here is we treat the figcaption as a caption. It's not an alternate text for the image though (which wouldn't be the same thing as the caption anyway). For image search, we'd use both captions and alt-text ( https://support.google.com/webmasters/answer/114016?hl=en ). Reusing the same text for both is probably not that useful.That said, instead of blindly following SEO best practices for any kind of image, I recommend working out how you think users would come to your site using visual searching. Don't think of Google Images as a "jpg search" but rather as a "visual search" tool. How might people be searching visually which aligns with your site's goals? Which kinds of queries would they be using? Which images on your site do you need to make accessible for that? I realize that's less "tech SEO" and more "marketing", but just how you can't just add all the keywords to a page as h1's to make it rank for everything, it also makes sense to be efficient and targeted when it comes to images.
Forum discussion at Reddit.