Namely, the following one
"All problems appeared once in the [American Mathematical] Monthly."
I remember reading it several years ago... When I first posed the question, I believed that I had read it somewhere in Krantz' Mathematical Apocrypha but according to Carlo Beenakker the quote is nowhere to be found in the said recollection of stories and anecdotes of Krantz. I gather that Erdős might have expressed it in relation with the history of what is nowadays known as the Sylvester-Gallai theorem but it'd be perfectly possible for me to be wrong in this respect too.
I thank you in advance for your replies.