It's an interesting explanation, and I too couldn't opine without further research.
I do have a comment regarding its feasibility, outside the chemistry- 1% (by weight, I'm assuming) of a 15 gram aluminum can is just 0.15 grams of manganese. I realize significant quantinties of catalyst (by mass) aren't necesary to affect large masses of reactants, however, geometry and mass flow plays a big role. A catalytic converter (on cars or stoves) utilizes a catalyst coating on a ceramic substrate matrix shape (boxes, honeycombs, whatever) designed to expose a lot of surface area of reactants to the catalyst. In your hypothesis, we have just the opposite. A cylindrical interface for the reactants and catalyst, which is almost the least possible surface area/volume ratio achievable, and a very dilute concentration of catalyst. And not all 0.15 grams of manganese (within the manganese dioxide) would reach the reactants as MnO[SUB]2[/SUB] is solid up to about 1000F, at which point it deomposes into its constituants and would loose its catalyzing property.
I guess the bottom line for me is that there is no substitue for burning clean and regular brushing.