3D printers - reprap and the limitations of opensource

I've often found the concept of 3D printers interesting, I've looked at building a reprap several times over the last few years.
When SeeMeCNC introduced the H1 I bought one, in the end the total cost wasn't significantly less than sourcing a Mendel, but the low intro price pulled me in.  The build was poorly documented, (something since fixed) and involved a lot of pieces, but it wasn't overly difficult. Overall I have been stunned by the quality of print that can be produced, it takes a lot of messing about to get there, but the quality is better than I would have imagined.
I've since built a mendelmax, and I've designed and started construction on a 3rd printer.
So it's all great then, well no not really, I mentioned above there is a lot of messing about.
I would bet most people not technically minded would give up before they got a decent quality print, and even technically minded people would struggle getting there.
the problem is the vast divergence in hardware and software, the totality of which makes plug and play impractical.
Basically a printer builder, builds the kit, tries to keep theing square and freely moving, wires the electronics (one of several variations) then downloads firmware to the electronics, doing so requires editing C header files in the firmware entering values corresponding to the particular combination of hardware happens to include, compiling and downloading to the board. You then calibrate the printer, so that actual movements correspond to requested ones.
The would be "reprapper" then downloads one of a few options to convert 3D models to gcode, Slic3r and Skeinforge being the most popular, they setup between 50 and literally 100's of parameters, to produce the gcode.
They load another piece of open source software to drive the printer, load the gcode model, home the printer and assuming everything is mechanically sound and calibrated you get a decent print out of the printer.
Reprap is founded on the idea of fast evolution, hence the biological naming "Darwin", "Mendel", Huxley" etc.
Much like most open source initiatives, and evolution in general the problem is you end up driving towards good enough, but good enough in the context of the community that is developing the product. The problem with that is that it limits the broader appeal.
I thing we're at a tipping point in the space as commercial concerns start to drive the technology towards the main stream something the free market is better suited towards than the open source movement, products like the UP and Cube are shunned by the reprap community, but for a price not much more than a self sourced Mendel, you get a ready to run printer that with minimal configuration can produce excellent prints.
So many of the problems with configuration, go away as soon as you have preassembled, known hardware, and commercial concerns exploit this, providing known working configurations for the hardware they provide.
It's an interesting parallel for open source software which I think has many of the same issues, it's targetted at the developers as opposed to trying to entice end users.
I am a big fan of the reprap community, I love the concept, but I'd recommend a prebuilt printer to most of my friends.