Great comments!
I think it might be worth noting at this point that DIY is rather a large topic area all on its own. To build a Cmoy doesn't take too much electronics experience other than knowing how to drill a few holes and wield a soldering iron. But designing gear something like AMB, Tanget, or Morsel does requires technical expertice that few DIY builders posess. Ti's Beta 22 design clearly rises to the level of "World Class" amplification, and it would be far beyond most peoples ability to troublshoot one properly if it broke.
So, even in DIY, it's worth highlighting the fact that there are DIYers who just know how to solder, some that can trouble shoot simply (ohming stuff out till they find the problems), those that are real technicians (who can trouble shoot with an o'scope and schematic), those that can do some simple designing (knows to lay out componants around op-amps), those that can do real electrical engineering (bias descrete transistors and/or tubes in various topologies), and those that can do real esoteric designs (far more complex than I can fit between these parethesis).
But that's the great thing about DIY, it's an activity with a broad spectrum of opportunities to learn. Start simple as a kid in middle school, and work your way up to really designing electronic circuits, and you could find yourself old, gray, and with a satisfied grin on your face because you've had a lifetime hobby of good clean fun.
I am very greatfull to have Joe W. on staff who is an outstanding audio engineer, and who has the technical chops to both make subtle design and performance improvements in a Jung Diamond Buffer circuit, and to balance all the factors needed to build a commercially viable amp. But you're right the he, Jamey, and I don't really spend our time pushing the cutting edge. For us it's more about delivering as much amp as we can at a variety of price points, and with a variety of features, all the while ensuring that the gear is rock solid reliable. In many ways, that's just as difficult as designing a bleeding edge amp. To get ultimate performance is complex and time consuming work, but to get performance at five different pricepoints is difficult and time consuming as well.
But I think it is fair to say that commercial amps are likely to be somewhat conservative designs, and that the real esoteric stuff does appear from DIY. But the great designs come from a very small group of DIYers: Ti, Pete Millett, Kevin Gilmore, etc. But these designs do end up in some commercial amps; Gilmores designs at HeadAmp for example. We've taken a DIY design and turned it into a commercial product---the Millett Hybrid---and we'll probably do so again as intreaging designs show up.
Someone mentioned the Cary 300SEI, I love this amp, but Pete Millett told me that the really good stuff is to be found elsewhere. The
European Triode Festival is one such place. Last year a few of these guys actually made their own tubes! I'm actually going to be attending this year as Pete and I consider the future of tube gear at HeadRoom. I guess I believe that great designs are great designs, and if a design is stable and well behaved it can be made into a commercial product.