One thing to add to this is that well-written apps built from the ground-up to take advantage of OSX technologies tend to get noticed and promoted in the Apple community which may have the potential to offset the smaller user base. Look at the exposure
Squirrel got when it won an Apple design award. If that app had been written for Windows, I venture to guess it wouldn't have done nearly as well.
Mac users are typically willing to spend a little more money for quality, as opposed to simply getting the cheapest thing available, so keep that in mind as well. That's not to say that you can charge 10% more for a port of a windows app. IMO, Mac users are more likely to pay more for an app that's well written and "feels" like it's been designed for them then for something that "feels" like a windows port.
If you design your app around the lowest common denominator in terms of being highly portable to multiple operating systems, you may loose out on the benefits of an enthusiastic OSX community. Returning to the Squirrel example, had that been designed for OSX, Linux, and XP/Vista many of the features that make it so appealing to Mac users would likely be gone (such as smart folders and spotlight-like searching). Instead you'd simply have a blah personal finance app that doesn't really compete well with giants like Quicken and Quickbooks.
Obviously this is a fairly specific example, and there are a lot of variables to think about when making these types of decisions; xyzeugene did a great job pointing many of them out. I just wanted to point out that it certainly seems possible that designing an app around a polished "Mac-centric" experience can make good business sense in some situations.
The more details you're willing to share with us, the more helpful the advice will likely be. You might want to give us an example of an app in the same ballpark as what you want to do that's already in existence, and you may get some helpful feedback based on that.