Thanks for calling attention to this interesting article. I was wondering whether the increasing use of computers (beyond numerical computation) might force the development of new notations that would aid in representing abstract or complex problems to machines. This would be similar to the evolution of notation in symbolic logic, which the article mentions.
We've all certainly had to deal with notation in writing for publication. Halmos talks about this in his pamphlet How to Wriite Mathematics (see the section Think About the Alphabet), available online by a simple search.
There are challenges that may be special to probability and statistics. For example, we might always want to represent random variables by capital letters (preferably toward the end of the alphabet) -- but we also want to write an estimator by putting a "hat" on top of a parameter. Similarly, we may want parameters to be distinctly Greek, but the binomial parameter is almost always written as a lower case p. And so on. Notation needs to be consistent, but also familiar. In writing, the choice of notation can be a bit frustrating, but it's worth the effort.
------------------------------
Jay Beder
Professor Emeritus
University of Wisconsin-Milwaukee
------------------------------