However, I do think the closest moral analog to running a casual Farmville type game is running a casino. If you're fine with running a casino, I'm not trying to stop you.
If you really want to find out if targeting these users to maximise profits is harmful or not, I think it's a good idea to ask these spenders, perhaps have them take a survey. If you don't want that, I'm pretty sure there's research out there on gaming and spending habits that can shed some light on the question.
[Edit: I see that you sell through iTunes connect, and I don't know if Apple permits you to have anonymous user surveys in you app. Even then, I'd guess that you wouldn't risk your highest paying customers reconsidering their gaming habits by asking them uncomfortable questions in a survey. Given that, I would think actively looking for published research papers would be a good idea.]
Targeting is your word, not mine. We cohort users by spending activity and learn from them in the same way a Macy's buyer will learn from their most loyal customers. My friend works as a buyer for children's clothes and they make the same business decisions I make about inventory and store layout.
We maximize for users who spend large amounts of money (incidentally none of ours have spent $10k, more like $100) because this and engagement data is the only data we have in determining customer loyalty. So we ask ourselves questions like "If this user found enough value out of this feature to spend $100 on it, we should make more like this".