It works well with a 1/4 birch plywood piece to replace the aluminum.  You'll notice the bumpers on the side to hold the glass in place.  I have a couple clips to keep the glass held down tightly.

With an extra laptop power supply attach via relay and flyback diode, it heats up reasonably fast.  Typically I turn on the printer  and load the file.  By the time I have the slicing done, the bed is up to 65 degrees or so.  I find that the bed is much more evenly heated.  I print with abs at 70 degrees.

If you compare the different solutions.  The resistance of the heating element is pretty critical.

Standard issue resistor - 3 ohms
Standard replicator pcb bed - 10 ohms
QU-BD Silicone Bed - 1.6 ohms

Please note the resistance was taken with care using a fairly new Fluke 87V.  For readings that low I'd be more comfortable with a milliohm meter, but don't have one close by.

Some interesting thing happen with the different beds.

The replicator uses a 24 V power supply.  I happen to have an extra old laptop power supply that seemed to do the trick for me.  I first tried a relay.  I need to put in a diode as a flyback as the relay clacked on and off.  Ideally a solid state relay would do the trick.

Using Ohms law I see the following  (P=V^2/R)

24V with 10 ohm bed yields  57.6 watts
12 V with 3 ohm resistor  48 watts
12 V with 1.6 ohm resistor 90 watts

I did the math from the reprap bed page.  They figure 120 Watts is ideal.  http://goo.gl/k5nEa

I think having a second power supply becomes pretty necessary.  The 7.5Amp draw of the silicone bed is a bit much for my taste.  It seems to me that the ideal way to go is to follow the PCB design, but rebuild it for 12V and 150mmx150mm.

I designed a bed, but got delayed on the production.  This is good, because I'm thinking an adjustment of the traces is necessary.

Is there any interest in a group buy on PCB beds with a second power supply input/ solid state relay board?