- There is an exploration of failure. Most corporate people don't like talking about, let alone mapping out, things that went wrong. Obviously finding people to talk about failure in detail is difficult.
- You can't be politically correct at the expense of accuracy when creating a simulation. Can you imagine making a simulation of a car that works the way the designers hoped a car would drive, instead of the way a car really drove? Of course not. Likewise, simulations that present idealized world views are quickly found out.
- The more content, the harder it is to translate to other languages. Many people act as if ease of translation of a course is a good thing, not a bad. Of course it takes less time and resources, but less content is less content. Some say, "simulations get into cultural issues and behavior, which makes them harder to localize." I say, "simulations get into cultural issues, so they actually work."
I think the best quote about simulations comes from outside the field:
Nineteenth-century mathematicians discovered to their discomfort that as the conceptual machinery of mathematics became more precise, it became more difficult.—David Berlinski, The Advent of the Algorithm