I take it that somewhere in there you have pullup or pulldown resistors to prevent floating inputs...? Because it sounds like you don't....
I don't have any external pull-up resistors in my circuit anywhere, but I was under the impression that if I set a pin to INPUT mode and explicitly set it to HIGH (which I am doing), the internal Arduino pull-ups would be activated per the documentation:
"There are also convenient 20K pullup resistors built into the Atmega chip that can be accessed from software."
(http://www.arduino.cc/en/Tutorial/DigitalPins)
Again, I never had this problem before I built the proto breakout board, and if I remove it from the system and plug the sensor leads directly into the board, the problem goes away. I may just go back to this approach because it works, but I was hoping to be able to keep the spring terminal blocks for convenience. And at this point, even if I do go back to direct connections, I'm extremely curious why on earth this is happening anyway.
I left the proto board plugged in all night, and when I came back this morning, I saw something very interesting: all of the touch combinations were registering as active, with an occasional fluctuation on one or two combinations. Usually, less than half of them register this way while the proto board is plugged in. This appears to be the result of leaving it plugged in and turned on for so long, and appears to be in line with the weird possibility of a charge buildup causing the problem. I pulled the proto board off the Arduino and all of the sensors immediately reverted to "not touching." It isn't anything on the Arduino board itself.
Just to test, I plugged the proto board back into the Arduino after about 15 minutes, and the random behavior that I'm used to started up again: somewhere between 1/3 and 1/2 of the touch combinations showed as active. After I left it in for a couple of minutes, it progressed to the point of nearly all of the combinations showing as active. This happened faster than I expected, and if it is a charge buildup of some kind, that indicates to me that the charge didn't have a chance to dissipate in 15 minutes.
The actual proto board that I bought is here: http://www.sparkfun.com/products/8811
One interesting thing about the board is that one side has square solder pads around each hole (as I expected), while the other side has a continuous metalized area (for full-board ground?) with small breaks around each hole to prevent shorting every hole with every other hole. If you've used one of these boards from SparkFun or a similar board, you know what I mean. Let me know if I'm not explaining correctly.
I have connected the two GND pins broken out from the Arduino board to this large "pad" on the opposite side of the board to see if it helps, but it didn't change the behavior at all.
I don't know a whole lot about the physical build of electronic components, but I know that a capacitor is made of layers of specific material placed close together. Is it at all remotely possible that the proto board construction is exhibiting capacitance here? The Arduino documentation says that pins in INPUT mode need only a tiny amount of current to change state. Do I need to add pull-up resistors beyond the ones in the microcontroller? I rapidly shift each I/O pin between INPUT and OUTPUT mode during general usage, so I don't know how well that would work.
I'm still confused. :-?