I actually do not think that DSLRs need to be a dying breed. But, after I heard that mirrors themselves were labor intensive to manufacture and that EVFs were fairly inexpensive to manufacture, I knew that this transition was inevitable. Right now I suspect R&D costs are high, but eventually, the manufacturing costs are going to be lower for mirrorless cameras. In a shrinking market, in only makes sense to transition over to a technically equivalent (plus or minus) but lower cost technology.

So, as much as people might say this is about evolution, and technology....really, it is mostly about manufacturing cost. There are some advances that benefit from removing the mirror such as fps would always be limited by how quickly the mirror could move, it does appear some lenses will benefit from the shorter flange distance, and the noise the camera makes when taking a picture. But, I think they could have figured out how to achieve many of the advances being touted by MILCs in DSLRs had they invested in that R&D. Such as getting much better AF point spread in FF DSLRs. And as for the FPS, the 1DXIII has a mirror capable of 16 fps. As for Eye-AF, the "deep learning" in the 1DXIII is real, and I suspect with enough points in a AF sensor, they may have even been able to do Eye-AF with a PDAF system. But why develop new tech when a competitor (Sony) is moving to a technology that will eventually give them a cost advantage? Plus, if you think about it, Canon was developing "mirrorless" technologies for liveview as well as the mirror/PDAF system. So now they are developing tech for a single AF system that is used across all their ecosystems (cinema, R, maybe even security and other markets).