With the greatest of respect, this thread would be a lot easier to read and understand if there was an actual example of the trade in place - then people would be able to give constructive feedback. As much as I love talking Greeks, the over-weight dependence on positive Greek1/negative Greek2/sideways-positive-with-sugar-on-top Greek3 doesn't make it easy to follow what is going on.
No correct answer to this, but you are. Overhedging your positive gamma is a good way to never make money. Take your model and sim hedging every few ticks and you will see the extraordinary amount of hedges you have to do in a day just to make back your theta. Almost impossible. Now compare that to hedging once at the end of the day (not suggesting this is the correct answer, just illustrative purposes.) I bet the later beats the former, and with a hell of of a lot less work and slippage.
Way back when I was on the Cboe floor I noticed successful traders paying more than theoretical value for certain low priced options. Also, these traders would not hedge all the theoretical deltas from these options. The options seemed to have a low price but would vary between stocks. I set out to define what this level was. Traders back then were secretive but I as able to get some data points from many of them. What I determined I came to call the 'Rip Value', called that because the traders would rip up the tickets they would usually give to their clerks to add to their positions to hedge. The Rip Value is the value where traders would start to pay more for options and also start to take deltas out of their positions. The Rip Value is about the theoretical value of a 45 day -25 delta put. ORATS uses the Rip for position management and also as a trigger to exit options in the backtesting platform.