@Richard_Pearse: I know you know all this, but you said it so neatly I’ll comment from your comments for context.
All Boeing’s narrowbodies from the 707 in 1958, through the now out-of-production 727 and 757, to the still on-going production 737 have exactly the same fuselage interior and exterior width and exactly the same seating abreast (3 and 3). If anything since about 1990 the aisle has gotten slightly narrower and the seats slightly wider (e.g. take 1.5" out of the aisle and distribute it as 1/4" per seat). But that was a one-time change and there’s no more room where that came from.
The only way to make seats wider is to reduce 6-abreast to 5 or 4.
AFAIK all A320-series are the same 3 and 3. The 320 is about 10" wider inside compared to the Boeings, which buys about 1-1/4" in each seat plus the same in the aisle. The difference is small but quite noticeable if you sit in both on the same day.
In terms of fore-and aft spacing, “seat pitch” is the industry unit of measure. But it’s a confusing unit of measure for consumers. Seat pitch refers to the distance from a point on one seat to the same point on the seat ahead or behind. So it’s the sum of passenger leg/torso room and seat thickness. Modern seats are 3-4" thinner front to back than were 1970s seats. So pitch can (and has) shrunk the same 3-4" without any difference in perceived personal space. To be sure, the very budget carriers did not stop there and have reduced leg/torso room even more from back then.
All exacerbated of course by Americans getting taller, wider, and thicker than they were in 1967. And given the skew distribution of people size vs SES, the increasing affordability of air travel post-deregulation and the free-fall in real pricing since has made more of the lower SES bigger folks more able to afford travel.
We run just two kinds of flights. 100% full turning standbys away, and significantly empty. The former flights are 95+% of all flights I work. We rarely leave with empty seats. But when we do, we’re not leaving with one or two empty. We’re leaving with 20 or 30 empty. Late last night on a massively delayed f***ed up flight I carried 20 people on an airplane with 170+ passenger seats.
The practical effect is our published “average load factor” is maybe 90% and on holidays more like 95%+. But it’s a barbell distribution with a huge fraction of flights 100% full, and a small fraction of rather empty. Our median flight is definitely 100% full. Like when applied to any skew distribution, the term “average” hide more than it reveals.