I don't know about pdf files, but text inside an IFrame is not indexed by Google. I believe Google does follow links inside IFrames.
Oddly, I believe, the <embed> element is deprecated in HTML4 but not in HTML5? The <object> element was preferred.
I could only guess that by virtue of its name Google would index "embedded" content.
There may be security issues/ differences on how browsers handle material in these two elements for content on and off site. Iframes may be alright for off-site content, but "embed" may have security issues for off-site content. Not at all sure about this.
I'd say the challenge is that a pdf file is NOT inherently a web file - that is, it's not html. This means that how it is handled will be specific to the browser and version. I would not suggest handling a pdf in either fashion.
For the last three years, G (Bing to a lesser extent) are reading PDFs okay, but to OP's original question, best I can offer is that G/B do read links... and a PDF is a link... in iframes. What value can result I cannot say. I only use PDF for a print specific structured presentation of an equally available HTML file, ie. I present both.