I don't think the above method is valid.
Note also that the teeth of the sprockets don't directly mesh because of the chain being involved.
I found this, and it makes sense as the math takes into account the involvement of a chain.
Interesting article. He much more eloquently, and thoroughly explained what I was trying to get at. Using the article as a guide, using 112 links and a 16 tooth sprocket (stock), it comes out like this.
The countershaft sprocket has 16, or (1*16) teeth. The drive chain has 112, or (7*16) links. The largest common factor for these two gears is 16, which is the common frequency. The gear mesh frequency of the counter shaft sprocket is 2,100 rpm x 16 teeth.
Hunting tooth frequency = [(gear mesh frequency) times (common frequency)] divided by [(#sprocket teeth) times (#chain links)]
Therefore, the C/S hunting tooth frequency is [(2,100 rpm x 16 teeth) times (16)] divided by [(16) times (112)] = 300 cycles per minute
Anybody feel free to check my math!
The article's worst case 38.2 and, having the countershaft sprocket and number of links equally divisible, we're way over that. I honestly think this is why Honda went with 114 teeth when, as we've found, 112 will work fine with stock gearing.
(The article went on to talk about smaller radius sprockets causing increased wear which I theorized about on another thread a long time ago. )