Within the context of physical display position
, as opposed to logical position within document
, I quite concur that the true probabilities of a given link being followed will indeed vary with position. This is an unavoidable consequence of the fact that visibility and actionability are directly postively correlated, resulting in non
algorithm is, of course, not the only such, but merely the most widely heard of; and, at least some of the others also enjoy USPTO granted protection. If another such contains claims re. such positional relevancy - and, I've a vague recollection of reading of one quite some time ago that does - then Google would need to include such factor with great care if it wished to guard agains the charge of infringement, unless it were absolutely confident that it could succesfully defend against such charge by way of asserting that any claims to newness were patently false owing their being obvious on their face to anyone skilled in the trade.
Of course, one could very well contest nearly all of Google's Patent claims on the very same ground!
In any case, it would seem to be the more likely that Google would simply avoid such issue by way of incorporating said positional factor into their SERP algorithm, which, being for all practical purposes a Trade Secret, is easily obfuscated, and appropriately alter the weight given to the results computed by their PR