(Clearing out the backlog of old drafts...)
This was pretty spot-on. $15 for a really well-written piece of complex client software is not even close to the right price level. People who want a high-quality email client that is not subsidized by any other revenue stream should be thinking along the lines of Photoshop, Maya, or Ableton Live. Yes, I'm serious. With the rise of webmail, native email clients have become a niche software category for expert users, and that is what feature-rich, niche software written for expert users costs, because the development costs cannot be amortized over a hundred million users who upgrade every couple of years.
In the not-too-distant future, you should expect to pay $500 or more for a really high-quality piece of client-side email software. Or else you should expect high-quality client-side email software to go away (with the exception of Thunderbird, which is being maintained by Mozilla for basically philanthropic reasons; we'll see how long that lasts).
Edit: I should add that various Free or semi-Free Software email clients are not really an exception. Their development is funded by a completely different model which permits sustainable, lower-cost, but also slower development. The lower development throughput of freeware email clients manifests itself in two different ways. Email software "for geeks" like mutt, gnus, and pine completely abandons the goal of conforming to expectations of modern UX. On the other hand, free email software like KMail and Evolution, which is allegedly targeted at non-geeks, tries to track modern UX trends but perpetually lags behind in features and polish. Note that "modern UX" is a notion that is always changing, and thus requires high development throughput to track, sometimes because the UI itself is complex (consider gmail's conversation view) and sometimes because there is also a challenging computational problem behind it (consider gmail's spam filtering).