Computational science remains uncommon in many computer science departments, particularly in universities without an engineering school. Is it not considered part of the standard computer science curriculum?
Yes. The Iron Man Draft of the Computer Science Curricula 2013 by the ACM/IEEE-CS Joint Task Force, contains Computational Science and Parallel and Distributed Computing as two of the 18 Knowledge Areas recommended as part of the standard curricula.
Furthermore, the draft contains the following specific recommendations:
The “Big Tent” view of CS. As CS expands to include more cross-disciplinary work and new programs of the form “Computational Biology,” “Computational Engineering,” and “Computational X” are developed, it is important to embrace an outward-looking view that sees CS as a discipline actively seeking to work with and integrate into other disciplines.
Computer Science curricula should be designed to prepare graduates for a variety of professions, attracting the full range of talent to the field. Computer Science impacts nearly every modern endeavour. CS2013 takes a broad view of the field that includes topics such as “computational-x” (e.g., computational finance or computational chemistry) and “x-informatics” (e.g., eco-informatics or bio-informatics). Well-rounded CS graduates will have a balance of theory and application, as described in Chapter 3: Characteristics of Graduates.
I like your own answer to your question. I wished this were how it was considered everywhere but I suspect this is not universally the case. In particular, the big-tent-view is not something departments often assume (no department, not CS and not Math). I created a computational science course using FEM software here at Texas A&M (MATH 676) which I'm teaching every other year. It was a bit difficult to get into the university catalog since the CS department did not want me to use the term "Computational Science" (it would step on their turf -- well, maybe so, but there is nothing even closely similar in their department; it's like staking a claim but then not using it).
My personal view is that Computational Science should be a set of courses (a couple of basic courses on numerical methods, a basic "software in computational science" course, and then computational-X courses) that are taught in an interdepartmental mode. But that's not what I see is happening today.