From what I see, your Canon SC1011 created a hybrid PDF with a picture of the sheet and an attempted OCR interpretation of it.
Draw is not a PDF reader. It a a tool to create drawings. It accepts PDFs as a collection of graphical shapes, in this case a huge number of “elementary” text boxes containing mainly one word, sometimes a few words.
Both the scanned image and the attempted OCR interpretation are displayed simultaneously. Unfortunately, the OCR guessed the used font was Helvetica (is this the real font or not?). But this font is not present on my Linux computer and is substituted for an equivalent looking one in a Sans family. The substituted font has not the exact same metrics as Helvetica. Consequently, though words start at the right position, there is some drift in rendering which increases with word length. It is not noticeable on short words and really apparent on the right part of long words.
A real PDF reader will probably choose to display only the original image or the OCR. But obviously, OCR failed on hand-written data.
Fix
Use a real PDF reader.Thanks very much for taking the time to read my post
And supply a knowledgeable solution
Ditio except , edge or chrome which isn’t a ideal reader
But thanks for the observation