In article ,
Chris Malcolm wrote:

There are two different approaches to designing a zoom. The first
approach is to design it to a minimum standard of performance
throughout the zoom range. The second approach is to design it to a
maximum performance standard at an important part of its range, and
simply do the best you can, without disturbing the best performance, at
more distant focal lengths than the best one.

This is true of designing anything, not just lenses.

A device has some operating envelope. It will perform better in some parts
of the envelope than in others. If I limit the allowable range of
operation to just the part of the envelope where it works best, I get a
device which is limited in functionality, but performs great over that
limited range. If I allow it to be used over a larger range, I still have
the good performance in the middle, but now you can complain, "But, it
sucks near the edge!"

So, which is the better design philosophy? Well, neither, really. They're
just different.

Why do we even have variable aperture, variable focus lenses at all? For
any given image, I'm only using one setting. Let's say I take a picture
with a fixed focal length lens (f = 85mm), focusing at 15.5 feet, and an
aperture of f/8. The fact that the lens is capable of focusing closer or
further away, or adjusting the aperture larger or smaller may make the lens
useful in other situations, but for that specific image, it's all just a
waste. If I told a lens designer to design me a fixed-everything lens for
just that exact combination of settings, he or she could probably come up
with a lens that produced a better image. The only problem is it would be
such a specialized lens, nobody would buy it.

Everything is a tradeoff.