Hi Germiston,
Your take on making sure the devices work before you throw the old std overboard is generally true, but if the tech has less moving parts (6 compared to 30) and the lasers are lower powered (lasers malfunctioning are #1 reason for failure) the engineers can come to the conclusion that it's "state of the art" and take the plunge.
as Suresh said, the key is in our waveguides and the coupling efficiencies we achieve and at wafer level to boot which makes everything else so much easier. (Of course easier and photonics integration are at odds with each other but at least Poet has something they can rally around, and as Suresh mentioned after the earnings 2weeks ago, it's soo good it will give Poet a "sustained competitive advantage")
SP