Note: In college I was an occasional columnist for Georgia Tech's student newspaper, the Technique. Unfortunately, over the course of several website redos, they lost authorship information and many old articles are listed under the name of the editor in chief. I assure you that I'm the actual author of all of these. :)
After a brief hiatus, I am back to writing my regular columns for the Technique. This one might be called the last column of my "Kazakhstan series," though I never thought of my columns like that.
While I feel like I ramble a bit, the thought process isn't so complicated. People have to make a tradeoff between money and convenience, and that's the quintessential question of product development. When you read advice for product development, one of the key points is always that your product should be a painkiller, not a vitamin: In other words, your product needs to solve an actual problem the target audience has, not just make things marginally better. If it isn't, you've basically created a luxury item. There's nothing wrong with that, mind you, but it completely changes your target audience and approach.
More critically, I think that when it comes to the issue of software development, creating "luxury items" doesn't work. I mean, sure, you could consider something like Adobe Photoshop a luxury item because you can get The Gimp for free, but Photoshop is the industry standard software that professionals rely on to do their work. It solves the pain these people have of photo and image editing more effectively than anything else.
In web applications, the situation is even more stark. There are a select few applications that actually cost money to use (and don't have a free version), and with a few exceptions these are targeted at businesses. If your target is the casual user—the one who might carry two cell phones because it works out to be cheaper—your product just may not be worth paying for.