Journey series mismatch errors... theory

News: Welcome to the Canadian Paper Money Forum.

We know for a fact that the two serial numbers printed on each Journey series note are printed with two different inks, one being invisible to infrared light and the other not invisible... e.g. see image in this recent thread...
http://www.cdnpapermoney.com/forum/index.php?topic=11364.0

Anyone who examines a bunch of Journey series notes will notice that the spacing and alignment between the two serial numbers varies considerably from note to note.  I simply cannot believe that both serial numbers are printed at the same time with the same machine.  So I theorize that each serial number is applied separately.  To support this theory, I would point to the high number of mismatch errors that have been reported for the Journey series.  Some of the more spectacular errors with 3, 4, or even 6 digits mismatched cannot be explained easily given the default assumption that both numbers are applied concurrently.  If they were, at the end of a run, the press operators would be able to see these errors and flag the ream of notes they just produced.  With each serial number applied separately, the press operators would be less likely to catch the error, and it would be up to quality control inspectors to find the errors on the actual notes.  Given the high number of notes produced by each printer, it's impossible for them to examine everything.

Any comments?  Has anyone proposed/debunked this theory before?

Here a scan of a picture taken from the Bank of Canada's book "The Art and Design of Canadian Banknotes".  We see clearly that there two numbering heads put side by side for each notes. 
{http://img217.imageshack.us/img217/2648/numberingapparatus.jpg:http://img217.imageshack.us/img217/2648/numberingapparatus.th.jpg}

But since this press can print many types of notes for differents needs and designs, the heads seems to be adjustable in relative height, so the the different height of the one serial from its neighbor on a note can be explained by a head whose tighning nuts has been more or less slacked and slipped during printing.

So there is no need to have two passes for printing the two numbers.  And to explain the two types of ink, it's only need to have two ink dispensers in which the printing headstake their respective ink before applying it on the notes...

Aaah, yes, I forgot that the printers are making notes for several different countries.  I also forgot that the numbering devices have the numbers backwards so that it makes the job of setting the numbers (and matching them up) more difficult.  Still, those errors with 3-6 mismatched numbers are hard to fathom.

Since there is no connection between the printing heads, they are completely independent in their numbering sequence and if a mishaps occurs affecting one of the 90 printing heads we find to print a complete sheet, it's only this defective head which will put the wrong serial and it can go for quite long before the operator catches the error and repair/readjust (or replace if necessary) the faulty numbering head.

I would believe that each numbering wheel in the printing head can move independently, in part for the operator can easily and rapidly set the serials  to the next sequence (between two printing runs).  However, I dont know if the stepping is completely mechanical (driven by the rotation of the drum) or electromechanical (driven with electric pulses synchronized with the passes of the sheets)...