That system would be the perfect candidate for fish line.

I kind of thought the same thing.  Just out of curiosity, what type of fishing line do you use.  A picture I saw you post in another thread didn't look like the typical stuff that comes to mind when I think of fishing line. 

That's how this system works as well.  It doesn't turn both motors to move horizontally or vertically like Core-X/Y & H-Bot.

Here is an overview of just the X/Y belt & pulley system.

I have no idea how practical this might be, or what kind of quality it would produce.  It's completely theoretical, since I just envisioned it last night. 

Unlike an H-Bot system, I believe that flexing shouldn’t be an issue with this one.  In addition, it wouldn't need to have 2 belts tensioned identically as in a Core-X/Y system (or the Solidoodle side y-belts).  Lastly, this system is different from both of those in that each axis is driven by a single motor.

So, what do you think of this X/Y system?  The possible issues I see are having an above average number of idlers, and an increased belt/cable length.  If the X motor cuts into the build area too much, it can be moved to the back by adding 2 more idlers (vertically), which would then force the need to use a round belt, fishing line, etc.

The real question will be, can you reduce the amount of friction on the drive system to the point that it surpasses what you gained in mass reduction of removing the extruder motor?

then "that shaft" which rolls on cross bar would have to be a rack and pinion instead. and all that possible slippage/slack and mechanical parts failure chances make for more issues. plus whos gonna restring that mess when needed lol

I know it's hard to tell from the picture, but "that shaft" doesn't roll on the cross bar.  It would be mounted on the print head carriage, basically taking place of the motor.  Here is a view from a different angle.  I also added the 123D file which is a much smaller file size than the .stl. 

And yes, it would add more places for slippage/slack, as you said.  But I know that kind of system for x movement works on our embroidery machines.  I only have to tighten the cable maybe once a year, and it pulls around a lot more weight than a simple print head.

I thought that adapting that system for the extruder as well, might be beneficial.  I have no idea if the benefit of mounting the motors remotely would outweigh any new issues that may arise by implementing it.  I was just throwing the idea out there as a possible alternative.

Yes, the top motor/system is that way on purpose so it can power the extruder drive pulley/shaft continuously.  While the print head is stationary, the motor would turn as usual to extrude filament. 

However, normal x/y movement will also rotate that drive shaft.  That's where the post processing script would be needed to counteract that rotation of the shaft to allow proper extruding.  The motors would be working in tandem similar to how the 3 motors on a delta machine do.

then "that shaft" which rolls on cross bar would have to be a rack and pinion instead. and all that possible slippage/slack and mechanical parts failure chances make for more issues. plus whos gonna restring that mess when needed lol

However, normal x/y movement will also rotate that drive shaft.  That's where the post processing script would be needed to counteract that rotation of the shaft to allow proper extruding.  The motors would be working in tandem similar to how the 3 motors on a delta machine do.

But the top motor doesn't do anything because there's no anchor point, it simply goes around in a loop, since the cable/belt is continuous.

The attached .zip file contains an .stl so you can see it in 3D, if you want.

For this to work, we would have to use a post processing script to adjust the extruder motor stepping to compensate for the rotation of the filament extruder drive pulley due to x & y movement.  I have no idea if such a system would be feasible or not, but maybe someone out there wants to do some experimenting.