Thanks for the DD, gottagetrich!
I enjoyed that video.
The one thing we have on our side is compatibility. And I mean that in more than one way. POET is implemented using the same infrastructure used to make CMOS. Taylor purposely engineered his process to make it possible for companies to make the switch to POET at minimal cost. Most of what has been proposed in these articles require exotic materials, completely new equipment and also lack many of the components that are needed to complete the process. AS in the video for instance, the research team was able to make an optical transistor, but there are no lasers, detectors, etc. IBM is probably one of the closest to a solution, but even they lack an integrated laser/detector.
The other compatibility issue for newer technologies is that they don't play nice with CMOS, while POET can be plugged in via waveguide or fibre to communicate with CMOS DSPs. The elegant aspect of POET here is that it can solve near term problems which amounts to sticking our foot in the door, and then implement a longer term solution to integrate the function of the CMOS role, and supplant it.
Taylor will be regarded as a technical genius for sure, but he also has business sense that other engineers lack. Shepherd emphasized this quality as a key differentiator between Taylor's approach and that of other tech developers who were ultimately responsible for destroying capital investment dollars.