mouser hit on something that occurred to me when I was just finishing the 1st page of this thread: I think that for some, perhaps many, who develop free/donation/open source software, there is a distinct disinterest in what it takes to run a normal business, perhaps most particularly with marketing and sales efforts. I can totally understand it, these are some of my least favorite activities. Now you could argue that free software without marketing is potentially as unlikely to succeed as commercial software without marketing.
What is potentially interesting to consider is the possibility that free is in itself a marketing factor. How you license your software affects how people think of, relate to, and talk about it. Free and open source have huge, independently motivated communities behind them that can become your evangelists without a lot of difficulty, provided you're offering an interesting product (and that may be almost the most fundamental requirement, interesting product). hsoft alludes to this affect on his products and I've seen it many times before. So in a sense people who choose free/open source as a model may be opting for "free" (in multiple senses) marketing as well, whether consciously or unconsciously, and may be doing so partly or largely out of a desire to avoid having to do explicit marketing. Given the actual potential impact of "free" on people's mindsets, this is actually a legitimate option (see the many large businesses today that base their profitable business models on "free", e.g. Google).
I also really like 40hz's idea of small devs banding together. It reminds me of the old GoD game publisher ("Gathering of Developers"), though that is not exactly a success story.
I'm also fairly certain I've seen some examples of things like that, though I can't recall specific links or product names unfortunately. I think the important point though is that it's vital to remove as many barriers and inconveniences as possible from the act of transferring money from user to developer. The idea of allowing SMS/mobile phone payments is very interesting for example.
Regarding users guiding software development/features, this has been discussed elsewhere on DC before, and it has its pluses and minuses. I would think many devs would be wary of having their dev priorities fully - or perhaps even largely - driven by user demand (although at the same time many would probably agree that their work is already partly or largely driven by a "filtered" personal sense of user need). As hsoft mentions, sometimes unglamorous stuff needs to be worked on. However I do think something as simple as allowing donating users to vote on features (and not allowing non-donating users to do so) could be a good approach. Just because something is voted on does not mean it's going to happen, and not attaching money directly to the vote means it carries less load and expectation.
I'm also glad to see 40hz brought up the "most people don't care" point and that this prompted mouser's previously mentioned idea of a fixed price up-front with a "show me other ways to pay for this" option. I had forgotten about this idea but I remain very curious about it and I'd really like to see someone try it on an already successful app (so we have a basis for comparison). I agree with 40hz, it's a chance to do some real experimentation and put actual numbers to our speculations and feelings on these issues. mouser, how about it? Screenshotcaptor maybe?
40hz also brings up the oft-discussed "manifesto" idea. I like the idea in concept, and there are various pieces of philosophy already scattered around the site, but I wonder if trying to distill and clarify would necessarily leave some of the things people love here behind. How do you reconcile a site that offers specific software for download and "sale" (donation), as well as support on that software, with a site that is a more general software-and-tech-advice community that is so much broader than the software the site provides itself? Well, DonationCoder, that's how! But is it really working as best it could? Would a "manifesto" focus and improve things? Are there parts of the site and its activities that really are unnecessary or just not successful enough to bother maintaining? Sad to say, that might be the review stuff actually, but I hold a special place in my heart for that.
Re: "humble indie bundle", after the success of the 1st one - if not even before that - I viewed it as somewhat tongue-in-cheek. Granted the circles I run in are much more aware of and into things like that, but I never heard any single game in those bundles described in anything but glowing terms. Many people I heard from who bought it already owned one or more games in the collection and had perhaps purchased them for even more than the average donation price of the HIB, but they bought again just to support it. One could say that was the exception to the rule but the success of the HIBs might indicate otherwise. Still I think the point that their success may simply be due to novelty and an unsaturated market is likely correct so it's not necessarily something to try to emulate, or if you do, try to get in early. It won't work forever I reckon.