Hello Swenger and welcome,
It isn't Monarch's fault If you write a PDF you can tell Acrobat to write a letter (a glyph, I think the geeks call it) at x pixels along and y pixels down from the top corner of the page. That's a different scale of complexity from reading a fixed width text file where the letter is in column x of line y. And Monarch does a pretty job of reading in PDF files and re-interpreting them as text files - I've handled scanned forms with multiple fonts, handwriting, Arabic and European characters, and recently a Visio diagram using Monarch, and got the data I needed.
You're doing the right thing by checking out monospace and freeform, and changing the scaling. If you have a regular report that you have several versions of, try to find the most stable scaling that gives you a usable text layout. Then, in Monarch, experiment with the floating trap option on your templates, which should give you the chance to handle data that drifts left and right a little.
If you need to handle multiple column regions in a PDF, please let me know, as there's a particular trick to deal with this that requires v10 or above.
Everything Olly has stated it correct. Working with PDFs can be cumbersome. A PDF is only a graphical representation of data stored within the PDF. The other issue is the amount of PDF engines and 3rd party writers that can produce PDFs. Sometimes the best results can also be to trap entire data lines and parse out what is needed via Monarch functions and logic.
If you have any questions the forums is an excellent place for help.
A floating Trap won't help me, because the column I am using to trap is constant, it is some of the other columns that are not aligned.
However the last suggestion will work. I grabbed the 2 columns as one and then used the Formulas TRIM and FIND( a space between the two columns) to split them.