The other day I was having fun chasing my dog around the house, when suddenly Rovio began deviating to the right side. Then I lost the right wheel altogether. Thinking of all the posts on this forum about defective wheel sensors, I figured that now after months of almost daily use, my bot now had the same problem. So I took Rovio apart, and to my surprise, it was not at all what I thought. If anyone is experiencing a problem like this, the first and easiest thing to do is to swap the sensors, from left to right. Still have the same problem with the same wheel not wanting to turn with a known good sensor? Then the problem isn't the sensors.
I then figured either the problem was inside the MCU unit (beyond my capabilities to fix) or a problem with the servo itself (possibly repairable depending on if it can be replaced). As it turned out, plugging in the opposite servo to the problem servo's MCU board socket revealed that the MCU was still functioning normally. Looked like a bad servo...
I then disassembled the servo itself. As it turns out, the electric motor itself had developed a problem. When a movement command was given, the motor would stall for a while and then spin a little weakly.
So, to make a long story short, a little cleaning of the armature's commutator and the brushes inside the electric motor, and Rovio is back to prowling the house again. Might be helpful if folks are aware that a dead wheel may in fact not be the fault of the sensors. At any rate, if the servo itself cannot be replaced, the motor itself looks like an extremely common electric motor that could be either purchased or found in many smaller electric RC cars.