This isn't to rehash the flame war that started but to list what everyone couldn't find/produce but either kept mentioning or denouncing.
The article was probably this one...
http://www.tomshardware.com/news/xbo...ood,10924.html
The direct link was...
http://www.rahulsood.com/2010/07/con...gainst-pc.html
Now notice first and foremost what Rahul says and who Rahul is. The only "debate" is that Rahul doesn't give up his source. Which frankly, happens all the time in journalism; so take it as you will. I'll also point out that this was posted over a year ago on Rahul's personal blog and was never removed. That actually lends some credence to some and means nothing to others.
Never the less, that's what people refer too.
As an additional note...
I've done a lot of looking over the last couple of days and spoken with both CS and Engineering friends of mine. The general position they both hold is that unless someone from the industry, with regards to a product they've already produced, comes forth and says specific, with regards to their programming, we actually cannot have any "valid" debate on this issue, on the grounds of console controllers.
As a brief summation I'll explain why.
A potentiometer is the device used to in analog controllers such as your PS3, 360, or even PC control device since the 80s. The problem is that most people (ie video game players) don't understand that these do not exactly measure in "degrees" of movement. So while the device has a "360 degree rotation" that isn't the number set that is sent over and read.
In the generic forms of Trigonometry, if we considered a left control stick on a modern console controller with relation to plane movement in a modern fps (Forward, Backward, Strafe Left, Strafe Right), we closely relate this to movement on the (X,Y) axises, which it closely is. However, if we consider the stick to be at rest at the Origin and we apply a 62 degree directional force, we assume the translation in the game is a N62E movement. The problem in short is that this is not guaranteed direct relation to what happens on screen.
There is a great chance that the extra 2 degrees (lost in programming translation) gets rounded down to something resembling 60degree movement for example. This is done because of how potentiometer works and is translated in code. So what we are potentially left with in our modern games is more of a 36 point position of movement on the left stick due to rounding rather than what is perceived as 360 degrees.
Again, it isn't that this is speculation persay as it is one possible method due to the hardware and software being worked with. As said before, unless anyone from the development side of CoD goes on record somewhere explaining how they program their movement values, we will never really know the answer, and even then, it probably differs from game to game depending on the team working on the project.
I'd really like to point out that this doesn't "prove" one is better than the other. This is just merely pointing out some information which usually gets left off. It's worth noting that none of this takes into account any information processed and translated by the right stick which in similar function acts as the equivalent of a Mouse. If anything, it only further points out that it is up to each development team on both sides to program around the hardware and decide just how retentive they want to be in their numerical values.