Yikes, those are some close calls! It feels like I am driving on a very narrow multi-lane road, and my side view mirrors are about to hit oncoming traffic. Tighter leading is one thing, but I think you have gone too far for this typeface. Could you find something with a larger x-height and shorter a/descenders?

I actually don't think those are too close at
all, and the leading isn't too tight overall
for that line measure (probably because the
x-height is so modest).

hhp

Morison, 1936 and 1967:

Nick, are you saying he changed his mind about tightness?
Because there are many other difference there that justify
the greater linespacing of the second setting.

hhp

Right.

The oldstyle had been set without extra leading, then when the neoclassical style came along in the late 18th century, extra leading was part of the look—hence different specs for Garamond and Baskerville.

There is also the matter of line length, which has a bearing on rivers and the regularity of word space size, against which the effect of extender-proximity “islands” has a relationship.

And times moves on, so one individual may get bored with a certain kind of setting — and it might also go out of fashion.

All of these considerations may have affected the change of specs between the two publications of Morison’s tract.

I can't see a change of logic or taste between those
two samples. The differences of line-length, x-height
and point size are plenty reason to have a much larger
gap in the second setting.

In fact if anything I would expect the 30s to be
much more generous with leading than the 60s.

hhp

I think that Morison's first try was better, so I suppose I would advise the OP not to worry about the "close calls". Not that an extra 1/4 point of leading would hurt anything, but in general, longer ascenders and descenders are to be preferred over more leading for the same x-height and the same linespacing.

Because the ascenders and descenders contain information, hence more of the printed page is actually doing work presenting the text to the eye.

In practice, of course, typefaces have for some time tended to have small ascenders and descenders, so that one can get a larger x-height for the same text density (assuming no change in how condensed the face is, reduce the linespacing by 5% and increase the x-height by 5%, and you have the same amount of text on the page).

And it's definitely true that x-height does an even better job of conveying information to the reader than ascenders and descenders, even if they do something and the leading does nothing.

So using a typeface like Corona with generous leading is the worst case; using a typeface like Cloister Light set tightly is better; but the best case - from the standpoint of efficiency, not aesthetics - is using a larger point size of a typeface like Corona, also set tightly.

The ascenders and descenders should be just long enough so that, even when the face is used without leading, the eye can easily follow the line without confusion. So that may mean a bit more than the newspaper faces offer.