Stanley, reading Kees' posts, I don't see any place he suggested that we should make do with subpar or defective tools. It seems to me he is saying that (1) actual width doesn't matter, because we size the tenon to the mortise, and (2) parallelism or uniformity is not necessary and may even be undesirable.

Anyway, this image may be of some interest to the discussion. It's a page from the Seaton toolchest book.

IMG_1400.JPG

I assume this is a nominal 3/8" chisel. We can see that it is trapezoidal in section, and that the width tapers pretty quickly from .400 to .350, and that no one cared about hitting precisely .375.

One of my favorite parts of the Seaton book is the discussion of the saws, and how the taper grinding is phenomenally accurate, despite the lack of precision machinery or even micrometers. It's obvious that the toolmakers of the time (1796) were capable of great precision; they presumably could have made things parallel or rectangular if they wanted to.