For example, the Saturn V used two different kinds of fuel: LOX with RP-1 and liquid hydrogen. This optimized performance for the 1st stage booster vs the upper stages. This increased the cost and complexity of the ground support. SpaceX uses only one kind of fuel for all stages. This reduces complexity and cost.
If you build a booster stage that is robust enough to return with only aerobreaking, it is going to weigh more and be more complex. You pay for that extra weight for every launch. Note that some of the structure is only used for re-entry and is dead weight on the way up. Breaking with the engines means they are used both on the way up and the way down.
As Musk points out in his presentation, fuel is the cheapest component of the launch system. Therefor it makes economic sense to use more fuel to land the launch stages, which are the expensive components.
The people at SpaceX are not dumb. They came up with a different solution because they framed the problem differently. Rockets are hard, and there is not a single best way to build them. There are a lot of projects that use vertical powered landing: McDonald-Douglas DC-X and Blue Origin New Shepard are examples and NASA funded various prototypes. Aerobreaking is not the only reasonable option.