I had previously done a little research into this...if I wasn't running iridiums in 2 of my 3 cars, I would give it a try.
First off, a gap that is too small means that the spark duration will be very quick and the spark will be thin and weak. If the gap is set too large, the ignition system will not manage a spark and a misfire will occur.
From what I understand, sidegapping maintains a decent gap but allows more mixture-to-spark exposure area.
Team sidegapping up with indexing and its supposed to provide a modest improvement in power and mpg.
First off, a gap that is too small means that the spark duration will be very quick and the spark will be thin and weak. If the gap is set too large, the ignition system will not manage a spark and a misfire will occur
I think if a gap is unreasonably small there could be problems. Although, I've heard that some racers set gaps at .010 or so for extra power. I wouldn't do this in a daily driver type car. The theory is that with a smaller gap it will take less voltage to bridge the gap which means the current will be higher thereby producing a hotter spark. The spark may be smaller but it will be stronger.
Horsepower is how hard you hit the wall, torque is how much of the wall you take with you.
Plasma physics is the kind of thing that makes super computers worth thier money. Below the speed of sound the air can disperse the heat of air or an object moving through it (wind chill factor) above the speed of sound, the moving object(s) continue to heat up as the speed increases because the air can not disipate the heat as fast as the friction of the object moving through it creates it. Within reason this limits the speed of the propigating burn within a cylinder to the speed of sound(a ping). I believe that the same kind of thing can be formed at low speeds when the air becomes superheated during an ignition(thermal limit is reached). The shape of the plasma cloud is key to it's ability to correctly burn the fuel. This is the idea behind BOSCH 4s, and those new Diamond Fire plugs. My guess is that since the air in a cylinder is moving so fast, and is so inconsistent in composition, that it's best at low RPMS for the air to be ignited several times. This is why the ignition modules on the market offer up to 25 sparks per stroke at low RPMs. Is it was better to have a single spark last the duration of the cylinder's power stroke, or have several ignitions per power stroke. Will one long spark produce a helpful plasma cloud, or just a fuzzy burn? Do I buy a $20 lottery ticket or 20 $1 tickets?
Read an article at another gas savers site, and they said that by decreasing the gap, the spark voltage was lowered and the amperage increased. This increased the temperature of the spark plug, and that this was what ingnited the fuel, and not the spark. I'm not sure about the ignition part, but increasing the temperature of the spark plug, could cause the conductors to burn up faster, fuse it in the cylinder, or keep it above the ignition temperature of the fuel/air mixture. The later could cause an ignition during the exhaust stroke or intake stroke and a backfire through either the exhaust, or intake manifold. This probably wouldn't occur until the plugs had been run up to full operating temperature, or during some high speed or performance run, when they reached a peak operating temperature, and then suddenly your intake manifold exploded.