If they're scalable they would have to be text layers or shape layers as they are vector graphics.

There should be a way to extract the text into a directly editable format.. perhaps not in text but still scalable nevertheless.

I don't know how effective a batch OCR using Acrobat would be on the larger-size TIFFs but that might bring it together too.

Just a few thoughts but if you're happy with the TIFFs then I guess it doesn't matter, lol.