If I bet on every horse in a race on the tab, I would need to spend $117 to return $100. That is why there markets add up to 117%. If you decrease that outlay to $100, you return $85.5. I know I am not wrong about this. I can only assume that you are talking along similar lines as on another post, where you are talking about betting patterns of other punters(trifecta thread), and punting to beat them as opposed to beating the bookies. Whilst making an interesting read and obviously worthy of some thought, I don't see how it applies here...
As far as standard deviation goes, I don't have the same type of setup as you so I really don't understand what you mean. I know that consistant horses are better permanent betting propositions, but really don't see how it ties in with anything I am saying.

My point is: If 30% of favourites win paying $2.9-$3, you lose 10-15% of your money.
To turn this into a profit, you either have to increase the dividend to greater than $3.34 or increase the strike rate to greater than 35%.

As the favourites are determined by thousands of punters, I don't beleive for a second that I can increase the strike rate, leaving me with aiming for a greater dividend.
If I were to study the form and pick who I thought was going to win, there is no guarantee that I am gaining an edge. Yes I might be winning, yes this might last for a long time, but there is no proof that you have an edge. The only proof is your opinion and the fact is you are just gambling still.

I am not saying that you cannot gain an edge like this, I am just saying that you cannot prove that you are...
Whereas, if you bet on every $2 favourite and could back it at $4(wishful thinking), well I think you all know my point...