This is a good paper, but I say that with a caveat: Why did they use pseudo-code when F# or C# could have been used and the simulation tested easily. In the real world I expect that when someone discusses something in code, I want real code not pseudo code, especially with all of the potential languages that could be used.
Oh well that is my only complaint, the paper clearly describes how to mix Incremental Computation with Parallel computation. Hopefully this will help pull the undergraduate classes out of the middle ages of computational science. This is the type paper that I would use in an advanced Senior class with master’s students in it. In the real world, be it model simulations, games, finance, there is a need to implement software that can manage incremental computation with parallel.
In WPF/XAML there is the concept of the Background worker that can used to trigger parallel processing while the end user is performing data entry or foreground may need to be implementing a high priority processing.
Take a look at the following article, it is worth your time, and if you can read code, you can definitely read this article.
On the other hand, if you write code, you might wonder why didn’t the authors mention things like the background asych process along with the support for parallel process in Windows?