I won't argue with that, haha.
Edited to add: Part of the reason I seem all over the place is because I'm thinking very long and hard about what it is I really want to do. I've been sitting here writing and thinking much longer than I would like to admit (hah), but I think this post is a big step in the right direction.
Yeah, it would be a dream come true if I could develop an open platform that others wanted to use. In reality it's probably something that only I will ever play with, but part of the fun is trying to come up with a design that satiates my desires while being practical and valuable to a broader audience.
In the simplest possible terms, what I want is a small physical computing device that is entirely self hosting, and it should be able to reasonably utilize standard keyboards, mice, and displays. The latter part is what appeals to me about the old microcomputers. You plunk it down and plug it into the TV you already have, and you're ready to go. Of course nowadays we have PCs, so we tend to have mice and keyboards and fancier displays lying around, so it should plug into those. The physical computing bit is just to specify that it should be easy to connect to "the outside world." Switches, sensors, whatever.
Now, I know that basically describes the Raspberry Pi, but I think there's a lot of room for improvement there. As a purist, and from my own (perhaps unique) perspective, I just don't like the fact that there are so many levels of abstraction. Firmware, an operating system kernel, a C library, a high level language interpreter, libraries in that language, and finally your own user code. Ridiculous! And then there's all the cruft and complexity around that, entirely unrelated to whatever it is you're trying to do. What a mess!
Of course, there are advantages to a platform that is basically a regular PC. Linux is ostensibly robust, and there is a huge wealth of already-written programs and libraries out there, plus you get support for all sorts of devices and programming languages. Yet... from what I have seen, and from my experience as a newbie playing with various development boards in the past, little linux devices can be overwhelming. Relative to a PC, they're slow and awkward to use, so they're not always a whole lot of fun to play with. If you have a project in mind, the complexity and breadth of the system can be off-putting, and finding help is often hard because of the vast variety of hardware in use. If you do finish your project, it can be difficult to share with others who are running different distributions or lack the packages you've installed or modifications you've made. I'll say it again: it's a mess.
So, what if I eschew linux in favor of a much simpler system; one that is more straightfoward and has fewer barriers to entry? I would lose the flexibility and breadth of software that you get with linux, but I think there are plenty of cases where it would be a worthwhile tradeoff.
That brings me to a point you've mentioned a couple times now: this is mostly a software issue. I guess don't really need to come up with my own board. Maybe I wouldn't be able to fully realize the potential of the RPi, but at $25, it's probably not a bad place to start. I haven't bought one yet though, and I have my FPGA board, so I'm going to see what it takes to run a soft core.