I've run my version of the XP Game - something I call the extended XP Game - with plenty of non-IT people, often in mixed classes.
It works just fine, I find the learning points just come, I ask questions like "How does that resemble your work?" or "How might you apply that to your work?"
I find 2 big differences: time estimation and "product owner" - both with the game and in the actual work.
It turns out that in many fields - conference organising, HR, accounts, sales - the "how long do you think this will take?" question is never asked. Time estimation, as software developers practice it, just does not exist. Rather most of the work is deadline driven. This seeds a great conversation about whether time estimation would be useful to them, whether the could do it, and how it works with deadlines.
And to the Customer/PO role which I have in my Extended XP Game, this is rear outside of software too. It appears that in many roles there is enough customer contact that the workers are knowledgeable about what customers want. Again this leads to a conversation about whether having someone on point for prioritising customer needs (and saying no) would be useful or a step back wards.