The arduino is an entry level great for exploration or teaching. But I prefer to teach Micorchip because they support education. I have a Microchip educational rep that provides me with micro controller and other materials. When I attempted to contact Atmel they did not return my email.
But one has to ask why arduino has been so wildly successful. The shields, the software, the hype?
My thinking is that the arduino simply makes it considerably easier to get from here to there. As an instructor I would love to have people start with a system like this and in time dump the training wheels. Honestly I do not much care for the boot loader approach as I would rather see people using debuggers. But in some cases the boot loader does make sense.
Perhaps we should learn from arduino's success and move ahead.
Atmel has nothing to do with the Arduino; they merely supply the AVR microcontroller (ie, ATMega8/168/328 etc) that others use. Did you try to contact anybody with the Arduino group? Those are the people you need to contact. Or - just build your own Arduinos; everything is completely open-source, you don't need to go to a singular company to do this kind of thing. You just need to get the parts, and build.
I am not sure what the complaint is about bootloaders; I understand the debugging issue you bring up, but you can't get that kind of thing in an open-source manner for the AVR anyhow, and you need a different kind of hardware setup (usually, when people have asked about this kind of thing at the Arduino forums, they are referred to use AVR Studio or such - but if they are asking about this, they are already fairly advanced; but still, it comes down to cost - a debugging environment isn't cheap for AVR, from my understanding - and certainly not open source).
The thing I like about the PIC is that it has a much greater level of support (which comes from its lower per-unit price, so everybody and their brother uses them, thus Microchip has a much greater demand than Atmel), which means more are produced, etc - Atmel struggles with having to batch up runs of the chips, and when they are run, they are bought out really quickly, and so you have this "wave-like" production shortage issue.
You can't recreate what the Arduino group has accomplished; the time to do that is long past (what would've been needed, likely, would have been some kind of push back in the early days - mid to late 1990s or so - to have Microchip provide support and knowledge to create open-source compilers and such, that could be run under Linux, on the Mac, and Windows). It was a synergistic thing that just "happened". It didn't happen for the PIC camp, for probably a ton of different reasons. You can't force this kind of thing; no one knows the secret formula - likely there isn't one, and everything is non-deterministic.