Interesting observation. One of the reasons I started thinking about this is because one of my section mates has a nice 19th century Italian bass, with a 38" string length. The bass would probably be classified as a 5/8 or small 3/4, but the ribs are above average in depth. The bass does not look or feel unusual.
I've also seen a lot of basses being made with smaller upper bouts and sloping shoulders that compensate with a slightly larger lower bout (Rabbaths Quinol, for example).
Thanks again Arnold. Of course, changing the string length will change our landmarks (where the "D" falls and the first octave "G", etc) but these can be compensated for in the design. Your mention of the plates is very interesting. I'm not sure I quite understand it. Would you mind elaborating?
The issue then seems to be the danger of losing the boom on the bottom end of the bass as you shorten the string length. Then, I guess, it becomes a catch-22 as to how much difference in sound you're willing to accept.
Do you think a 38" string length really needs a higher gauge string? String manufacturers have been getting pretty good at making smaller gauge steel strings, with higher tension in the last 20 years. Which was part of my original query. Does the fact that we now have higher tension steel strings, with smaller gauges, help minimize the sound we might lose when you lower the string length from 42" to 38"?
The more years I play and the more basses I hear, the more I feel that it is all about the "box". There is such a wide range of sizes and styles and bigger doesn't always mean better. It is more about "quality" of sound.