EDIT: I should add that if someone legitimately can't use the app, I would give a refund. I'm just pointing out that unlike SaaS apps that have a policies that provide refunds regardless of the reason, I would rather work with you to solve the problem, because 99.9% of issues are solvable.
If someone feels they can't use your product - for any reason - it's going to feel unfair to them that they can't get a refund, no matter who's fault it is.
Note that the top reply to your thread is a dissatisfied ex-customer who felt obliged to post about his bad experience (fair or otherwise), and I bet that may have almost immediately cost you a few sales.
I'm not saying give everyone a refund who asks for one, but just saying "it's not a SaaS product and I can't turn it off" might not be the best rationale (and certainly won't impress any customers who are legitimately unable to use your product).
I would err on the side of trusting your customers, and accept that some might rip you off but most won't, and you'll partially salvage a bad product experience with some good customer service.
We did this with my VERY last product working at Humankind (Bullet Viral Traffic). It was a WordPress plug in, and you bought site licenses for it 1 site, 10 site and unlimited.
Every function call did a "phone home" to validate the license. When it came back unverified, we would only print an error message to the screen.
Can this be subverted, yes, but 99% of the time it wasn't, and, honestly, it was just a lead-in to training.
I realize this is your bread and butter, but maybe work training in on the back end for a little more money. Lead them in with this outrageously low price, then stun them with offering a recorded seminar/webinar for only a couple hundred bucks.
My company dropped $20k on training for InfusionSoft AFTER dropping $1200/mo just for the service. There is more money to be had in training than the product (99% sure).