On Monday, 24 September 2018 at 01:32:38 UTC, Walter Bright wrote:
D the language is well suited to the development of Unicode apps. D source code is another matter.
But in the article you specifically talk about the use of Unicode in the context of source code instead of apps:
"With the D programming language, we continuously run up against the problem that ASCII has reached its expressivity limits."
"There are the chevrons « and » which serve as another set of brackets to lighten the overburdened ambiguities of ( ). There are the dot-product and cross-product characters · and × which would make lovely infix operator tokens for math libraries. The greek letters would be great for math variable names."