[Quoth]

PDF is a sort of envelope with page size(s). It's primarily designed to proof and print. It can commonly have images of each page, either from a scan of a physical document or deliberately rather than text. Some are also only intended to be rendered on a screen or printed.

So even if the Kindle does try to index PDFs many would fail. I have no idea if it does attempt it. Most PDFs are actually unsuitable for 6" Kindles, especially the Basic 167 dpi model.

Ebooks are a kind of zip file with internal index, HTML file, font files and CSS files. They are thus easier to index.

Some PDFs have an OCR layer (for search), even if image based and in theory that could be used for indexing. PDFs are electronic documents, not ebooks.