I'm suggesting that some of the methods that SpaceX has employed to reduce costs of their rocket in terms of applying mass production techniques and treating the manufacturing of rockets more like how automobiles are manufactured on an assembly line has made a huge impact in terms of the cost of a launch. They are currently manufacturing more than a couple Merlin 1-D engines each week and plan to ramp up that production rate to even higher levels. They have also streamlined a number of things in the vehicle design to drop prices considerably including using consumer grade electronics instead of mil-spec equipment (using redundancy instead to achieve higher reliability) and several other innovations to really drop costs that haven't been used earlier.
That is the cake I'm talking about which other companies haven't been able to achieve... for various reasons. The Merlin 1-D engines aren't the highest performing engines and definitely have some strong limitations, but they are very cheap to manufacture. The same goes for the body of the rocket and other parts too. The vertical integration of SpaceX has also helped in terms of keeping the supply chain tight and keeping costs under control.
I also question how much actual savings will happen with reuse as there are definitely fixed costs that really limit how much it can reduce costs. 1st stage reuse at best only saves about half of the cost.... when done over the course of nearly 20-30 launches for amortization and assuming even low fixed costs. That still is useful and can make SpaceX very competitive, but it isn't nearly the earth shattering cost reduction that some are suggesting. SpaceX isn't even talking much any more about 2nd stage reuse, and all of the contracts using the Dragon spacecraft currently require a new capsule on every launch.
In other words, SpaceX really can't be depending on reuse for profit and instead must depend on other ways to cut costs in order to survive as a company while charging so little to the end customers. Admittedly, SpaceX officials have quoted a price point of $7 million per launch of the Falcon 9 to deliver 10 metric tons to LEO as something they are aiming at (mentioned at a satellite conference in Indonesia last year with commentary by other launch providers simply saying SpaceX is quoting nonsense). That is about 1/10th of the price currently, but I would assume that includes more than just reuse savings.