I'm sorry, what is it you're trying to avoid?
I'm not sure I know exactly how to explain, but I'll try.
When I cut the tail lights out of my 1972 Ford Maverick and modded a set of junkyard 1969 Lincoln Cougar tail lights to fit in the holes, I was pursuing a goal that seemed likely to yield something interesting, unique, and cheap. It was DIY in the hands-on and risky sense; it might have failed miserably and looked bad, or not worked, and left me with a sense of regret. On the other hand, since it did work, and got lots of positive comments from people, it gave me great joy.
They also made fiberglas kits at the time to modify how your car looked. Sure, it was DIY, but it had been done, it was pretty much guaranteed to work and to look good. It was more about your skills with bondo and sanding and painting than your intuition and inventiveness. The goal was not just to to DIY, but to do it just like the instructions called for, using parts which were guaranteed to fit.
I have been looking at the various speaker-building websites for a long time, because certain speaker concepts fascinate me; full range drivers and open enclosures and folded horns and so on. And this takes me into learning more about the history of these things; where did they come from? What were they like originally? So I have an historical curiosity, not just a desire to build a set of speakers that are pretty much guaranteed to sound very good, if I can cut wood products and follow instructions and purchase the right bits.
It's not the speakers so much as the road to getting there. Speakers that sound good are very cool; I love to listen to them. And perhaps one day I'll build a set from plans that are community works like those found on the websites you mentioned, and I'll buy the right drivers and the right crossovers and spend time making all my carpentry and laminating skills work well, and I'll have something incredible-sounding as a result. But that's really not what I want at the moment.
I like the old Popular Mechanics and Popular Science articles because they're imprecise; they're a starting place. No speaker parameters, no anechoic testing, no computer modeling. Plenty of room for error, room to hack and make changes based on gut feelings and hare-brained ideas and just play. Throw in cheap drivers, change 'em out. Turn the angle on a board, make it shorter or longer or thicker or whatnot. Experiment with damping materials and construction materials and drivers and wire and crossovers (for multi-drivers, etc).
What I'm looking for is not nearly as likely to produce something known to be good. Perhaps that's the point for me. Does that make any sense?