Activision wants you (as of this writing) to prepay shipping and insure the drums whereupon they will be tested and then replaced. If you want to wait 4-6 weeks (estimated) to get them back.
I'm not sure whether to thank them for the inspiration or vilify them for the awesome customer service and manufacturing but that is neither here nor there. On with the show!
UPDATE! (11/04/08): I have determined that the multi-hit problem was due to a flaw in the plan. The supports under the drum heads should not have been attached to the main chassis but rather to the drum heads themselves. Towards that end there will be a GH:WT Drum Repair Redux soon.
Step 1: Baseline Your Kit
I considered the performance of my yellow cymbal to be ideal: easily registered both hard and soft hits in the center and readily picked up hits to the far edges as light hits. It was also very easy to get a roll-type action with minimal force.
All other pickups were far from ideal. Congratulations to the yellow cymbal assemblers wherever you may be, job well done.
I decided that I wanted all areas of the pad to register hits if possible, or at least light ones on the edges. I didn't want a false-positive type of hit to register, however, a gentle accidental contact should be ignored.
Decide what you want performance-wise, test to find out how far from that you are, and let's dive in.
Step 2: Repair Solder Joints
Take the drums apart, scrape the sealant off the solder joints, and start over. If yours are like mine the red drum is all but non-functional. I touched the wire and it just fell out. I believe this is partially due to wire-routing. There was quite a bit on tension on the wire set for the red drum head.
If you pull gently on the wires while watching the sealant carefully you can tell which wires definitely need repaired. Any movement indicates the need for repair. I did them all, 6 contacts is no big deal.
Step 3: Cymbal Fix
The main problem here is the piezo element was not fully seated in the intended location. It appeared to have some sort of adhesive that was locking it in the place, albeit the "wrong" place. I decided rather than risk removing it and having to rework the mounting location or the element itself I would simply press it into place.
Because it had so play in the mounting the strike vibrations weren't fully transmitting to the element. I removed the cover, stacked a bit of foamcore (or cardboard, whatever you have) so that the cove just barely didn't seat.
After screwing down the cover, the foamcore added enough pressure to the element that it contacted perfectly and registered nearly exactly as well as the yellow cymbal.
(If I can edit this later I will get pictures. First Instructable so...)
Step 4: Drums Better...but Not Great
Unfortunately there was nothing (in relative proximity) to press against like with the cymbal. Fabrication was in order. I used some flat, thin, kinda springy steel and fabricated 3 spring plates.
I mounted another bit of foamcoare to the apex of the spring, added a spot of tape to prevent shorts across the spring, and mounted the assembly below the piezo element.
Step 5: Drums Are Awesome...too Awesome
When I first finished the assemblies in the previous step I placed the piezos in the drum head(s) under pressure. It was light contact but it was there and it was immediate.
Under testing things were crazy. A cymbal strike would initiate contact on all drum heads, as would contact on any other drum head. A good job gone bad.
It turns out I needed to have a bit of open space between the bumper and the piezo. A bit of adjustment on the springy bit and a slightly thinner chunk of foamcore and we're fully rocking now.
Step 6: Final Thoughts
It could be fixed readily at the manufacturing level, I think, but that begs the question of units already produced. The bumper-type units I fabbed could be easily made via injection molding with a bit of a slit clip at either end to avoid fasteners and drilling. Attach a sturdy rubber bumper to the apex and you're in business. Or pull the piezos themselves and remount.
There is the issue of crap soldering and wire management that would still need to be addressed. I submit that using tape instead of the sharp bends would be step one, better assembly of the elements and harnesses being step two.
Hope this helps some folks, If I can I will edit and add more pictures a little bit later. Happy warranty voiding! We makers don't need you anyway.