If what is described in Mav's post is the main point you got, I think you would be for a big surprise. The devil is in the details. For instance, when you scalp you remove from the total gamma's you may have dearly paid for , and if the stock continues its march you lose the gamma you initially bought for the larger initial size, and you may get as total reward less than the total premium you paid because of the reduced size due to the gamma scalp done at the wrong time and price. You need to design much more subtle quantifiable criteria to know if and when to scalp. It require a deeper understanding of options, but at least these criteria exist. They are not based on feeling or experience, but on numbers.
You can think of it as buying a number of straddles which you sell as baby straddles. Baby straddles need to be sold at particular prices and times, and subject to what happens next. The timing is even more demanding in scalping, but somehow it may not look so because people "make money" when they scalp.