Lions still have a lot of errors in their tax returns.   It not a bug in pdfmerge but the complexity of the tax code.  It might be smarter to just use paid tax software.  Most animals have all their investments in tax advantaged accounts & all the rest in a house, so they're not dealing with the problem areas.  Lions never bought a house so their tax situation is upside down.

----------------------------------------------------------------------------------------------------

The killer limitation is there's no real way to move cells around after they're used in a PDF.  Since spreadsheets are shared between multiple forms, every cell move would require remembering every single form which used it & reapplying the same cell shift commands.

The easiest solution would be Goog supporting aliases for cells.  There is a way to make an alias for a cell with right click -> more cell actions -> define named range   The named range follows the cell when it's moved.

All the named ranges are accessed in the top left.

The named range can be used inside PDFmerge to access the data, but it's really cumbersome to copy.  It can't have any spaces.  There's no way to get a link to the named range like there is for a cell.  The only way is to highlight the name in the menu.

Then copy it & paste it into the URL inside PDFmerge.  Maybe if the data is important enough, this can be justified.  There might be a way to automatically create a named range for every cell in a form & replace all the cells with the named ranges in every form.  PDFmerge has a replace cell function which can replace a single cell with a named range.

Getting back to moving cells around, there could be a command for pdfmerge to search for all of its values which don't match the values in the spreadsheet in the neighboring cells, using an expanding search.  The leading mismatch would be a value in the form linking to a blank cell in the spreadsheet.  That could at least automate the cell movements in a single form.  The most common case seems to be for cells to get shifted around & new data to be added before the next sync, but the change doesn't affect the previous values.  There could be a button which performs the search for moved cells outside of the sync operation.

------------------------------------------------------------------------------------------------------------------------------------------

Another pain point is the challenge of opening the right localhost:80xx address in the browser for the right filename.  The best option might be multiple instances of pdfmerge in a single 8080 server with the filenames in the URL.  The easiest option might be a command line tool which shows all the pdfmerge addresses & what file each one is editing.  It would run ps xa|grep pdfmerge.py to get all the PIDs with filenames, then  netstat -tulnp to show what ports they're bound to.

The web app should have a redirect page with links to all of its own instances with the filenames.

Having to use a web app with goog's spreadsheet API is the source of a great many limitations.  A ground up spreadsheet program & native form editor would be ideal.  It would entail importing all the data from goog sheets & open office. 

-----------------------------------------------------------------------------------------------------------------------------------------

The syncing process with goog sheets is glitchy & slow at best.  Sometimes it needs a few tries to get the latest values.  It really needs to download the entire spreadsheet with a single GET instead of shooting a new GET for every cell.  This means a form can only reference 1 spreadsheet, but this has been the only use case.

An idea gaining favor is going back to open office & using an XLS to CSV converter to sync with the forms.  There's no way to copy the location of a cell in open office like there is in goog sheets.  You'd have to enter filename:B20 manually.