For developing new forms of printers.

There has been speculation about making alternate forms of robot to move the print head/bed. I have been considering using servomotors in a delta robot, as well as in an arm, eliminating the need for all the expensive stepper controller hardware, as well as making the motors themselves cheaper (servoes are cheaper then steppers, not to mention it will now get real-time feedback about its position. It will never back up again...) not to mention that with the right coding, it could eliminate the need for support material ( I heard mendels can print upside down. Why not just rotate the print head to paralell with the ground?) Has there been any testing on this subject, or is it all new ground out here on the bleeding edge?

Motion control really isn't the bleeding edge, industrial positioning systems have been at work for many years. Easy and relatively cheap access to them is newish, but again, nothing really groundbreaking.
Servo control systems are rarely cheaper than steppers, especially accurate servo control. "Real time" position feedback is not a trivial thing, and complications arise when the mechanical structure of the machine isn't as rigid as it can be. In most commercial and just about all hobby level servo based motion controls you mount an encoder on the motor shaft directly, as even the slop in ballscrews, end support bearings and linear guides is enough to throw off a servo control loop. Some (and I stress some, as the machine running behind me certainly doesn't have them) use linear scales on the table in addition to encoders on the motor shaft, but they are there as an aid to the control, and not directly in the servo loop.
Yes, you can get a cheaper DC motor than a stepper motor. Yes you can throw together an H bridge pretty cheap. To get an encoder that will match the steps per rev of a 1.8 deg stepper motor being microstepped is $20-30. And then you have to tune the algorithm to avoid overshoot, and damp correctly. There's code out there, but every mechanical system needs to be tuned to it's own parameters. If you know what you are doing, you can rig together something for just about the cost of a stepper system (system performance between the two being roughly equal) but most likely you won't save any money.
A properly sized stepper system is plug and play. If you size it right and power it right you won't miss steps, and there is very little tuning.
I'm not trying to rain on your parade, but there is lots knowledge out there for those willing to learn it. Many people, myself included have built robots like this. I have a 4 axis machine that routinely cuts parts within .001" and better. Do some research, read a little bit on CNCzone, there are other similar sites out there.

What I meant was that for the delta robot, there's 3 servos and one extruder motor. Having a BFB Rapman, I'm not considering the extruder motor now, but there are potentiometers in the servos, and they can be dialed exactly to where they need to go. In addition, a standard, run of the mill Arduino duemilnove can handle the 3 servos unmodded, maybe even off of batteries (though I won't be portable-printing any time soon, it is a thing one may want to work for). Not to mention, it will not need calibrating, or, if it does, very minimal calibration, due to the servomotors having their own potentiometer-based encoders. Check it: [letsmakerobots.com] Slap a platform on top of that thing, a stationary extruder above it, and have a nearly fully printable printer before you can say,"Holy delta robot, Batman!"

I'm somewhat familiar with hobby servos, I've used them before in many projects, including bipedal walking robots. To begin with positional accuracy of those units is not going to cut it for motion control of the sort needed to make a printer. Stock servos do not give positional feedback, and using the potentiometer to do so is going to leave much to be desired. A potentiometer is NOT an encoder, and does not function anything like one. In a hobby servo it's simply a voltage divider.
I'm also pretty familiar with motion control on the scale used for milling machines, and pick and place robots. Delta robots of the kind that move this 3d printer with positional accuracy are not as trivial as you make it out to be.
It's not like a delta arm is a bad idea, but hobby servos are not going to cut it for this type of work. Servo positioning systems that would work will cost significantly more than what you are talking about.