Unfortunately I think this is too big of a project for WA Team to be occupied with. Sorry it's a "NO" from me.
Not a terrible idea, but this would be a ton of work to implement; I'd prefer that the dev time goes elsewhere, as I think the convenience of being able to do this would ultimately be a minor benefit in comparison to the cost of implementation.
That would be very complicated to implement. Not only would WA need to parse the different file formats correctly, it would also have to identify headings, tables etc. which could be in ANY form, depending on who created the PDF/DOC/etc. and how they chose to format and layout their text. A human can understand "oh, several lines in a different font size and color, those must be the headings" - a computer will have a very hard time with that. That's almost a scientific field in its own right.
PDFs are less like text and more like images, but Doc files shouldn't be too much trouble. Although I'd rather the developers spent time on something else
Depending on how the files are setup, you could try CTRL + C and then CTRL + SHIFT + V to paste without formatting. Pictures you'll have to upload and reference one by one. Conversion will take a while but you'll get there. I am afraid this would be a LOT of work to implement.
This would be a massive challenge. WA's formatting is very different to that of a PDF. Export things as text, upload your images to WA (have you just been pasting and expecting it to work..?), import tables using the existing table import, and work on things with the existing flow - it'll make for a better storage structure, too.
Although I understand your issue - this is not a suggestion Hellfist, it is grossly generalised and non-actionable.