As a physics undergraduate and then a doctoral student I remember that one of Moseley's original graphs of the relationship between atomic number and X-ray frequency was displayed on the main staircase of the building I worked in. The department was rightly very proud of it. One of the things which struck me about it was how large it was - somewhere between A1 and A0 as I recall !
I'm just about old enough to remember physical graph paper being used 'in anger' to interpret experimental results. Our stores kept quite a range. As well as the usual linear stuff (which we had printed from our own plates, to save money) you could also get various types of Log-Lin and Log-Log if you needed to determine power-law relations between data sets. If your data had very low error bars then bigger bits of paper could actually give you higher precision !