So many fallacies in such a short post. Let's take the largest one first.. The price consumption curve is, as the name implies, a curve. Scalpers don't supply the entire market: they supply only the upper portion of that curve: those willing to pay the highest surcharge. The people here in the forums grousing that "they'll never pay scalper prices" are the lower end of that same curve. When a manufacturer raises prices, they affect the entire curve. A scalper's price affects only a small fraction of that curve.
Let's illustrate with something more concrete. Sony offere 5M units of a console at $500/each. It sells out, meaning there is a certain level of of unsatisfied demand. Some of those unfilled customers are only willing to pay MSRP or slightly higher, whereas others will pay double or more. Now, had Sony introduced the console at $650, then at least some of those who purchased at $500 would no longer buy. Unsatisfied demand is ineluctably lower, and thus the price point required to balance it is lower as well. The price drops - and because elasticity at this point is higher -- it drops more dramatically. Do you think those prices on Ebay come out of thin air? They float to a degree, yes -- but over time they float to the level that exactly balances scalper supply with that unsatisfied demand. A higher MSRP raises the ability of scalpers to buy (increasing their supply), and lowers the demand non-scalpers have for their product as well.
There's a third factor as well. Scalpers have ancillary costs -- Ebay commission costs, extra shipping costs (you have that effect exactly backwards in your analysis) -- as well as a certain degree of risk in any transaction. Selling a $500 console for $900 allows them to easily pay those costs, and still balance the risk. But selling a $650 console at $750 is a much dicier proposition.
In summary, the notion that, if scalpers are selling at a 50% markup implies that manufacturers would have to sell at the same price to entirely eliminate scalping is utterly incorrect.