I don't know whether I should concentrate on user-friendly programs or my own crazy preference for minimalistic but keyboard-shortcut-filled programs. I like customizing my laptop to be super-efficient, but it doesn't help anyone except me and other extreme nerds since usability has an inverse relationship with the number of shortcuts and jury-rigging present.
For example, I'm using mocp for playing mp3s and wrote a perl script that deletes the currently played song on the pressing of a hotkey. Who would use this? It's not even written robustly. Another example: I removed some pauses in w3m by editing a couple lines in its source code. I don't feel like sending a patch because maybe others find those pauses useful. I didn't make it an option--I just hardcoded it. I guess I'm not confident enough in my coding to post anything, but maybe later. Plus, I don't even know how to use diff/patch. The docs are tremendous. Arg.
The benefit of working with user-friendly apps is that I can provide support for people who want to try Linux. If I learn Gnome in and out, I can evangelize GNU so much more easily since I wouldn't be punishing my computer semiliterate friends by forcing the command line (or some obscure window manager) on them. Not everyone wants to spend 20 hours of research to produce a shortcut that trims a task by a mere 2 seconds. Not everyone wants a lightning fast browser that has trouble displaying Flash and RealMedia, regardless of how evil proprietary formats are or how fat and slow common browsers are.
If I use common apps, I don't get to feel fast and efficient, but what is efficiency anyway? Twenty hours to save two seconds? I really need to stop trying to be efficient since it's making me incredibly unproductive. I had just used Windows to create a website in about 3 hours without too much thought or hassle. If I had done it the Right Way, it would've taken two days probably. I have to strike a balance between meta-productivity ("efficiency") and measurable, invoiceable productivity. Currently, I spend way too much time getting ready for tomorrow's projects by working on new skills, technologies, and processes that I don't get today's projects done.