I've been thinking about this question some, lately.
There are at least two kinds of realism that I think of in this context.
The first is plain and simply the probability of playing the game. If I don't currently have time to play 1 hour games, I probably shouldn't be buying a lot of 4 hr games. If I don't have a game group TI4 may be a bad choice for me. If I don't like heavy games, maybe I shouldn't bebe buying a lot of COIN games.
The second part of this question about how real you are on your wishlist has to do with the likelihood of acquiring the said game. I really really want the black box edition of Glory to Rome. But, with its scarcity I don't think I will ever buy it. A lot of the Chip Theory Games catalog fits this category for me. If I'm spending $130-150 on a game, there will probably always be something higher on my wish list.
So, how do you keep your wishlist? Is it pretty realist? Or is it more idealistic?
For me, I think it is mostly realistic. That does not mean that I treat my wishlist as a shopping list. Not does it mean that I don't buy games that aren't on my list. But, I do think it is mostly, not completely, realistic.
This does remind me of something that I wish we could do on our BGA wishlists. On BGG you can weight your wishlist by your level of interest. I do wish that we could do something similar to that here on BGA.