In future I would like to see some OCR engine integrated to pdfcreator.

For example tesseract
http://code.google.com/p/tesseract-ocr/

Probably tesseract already has mature code that could easily be integrated to pdfcreator - unfortunately I dont know so much of programming.