I will add that what POET brings to the table here is the bandwidth.
For a device like the Apple Watch to ever collect glucose measurements (which is a game changer), it will need to do so by shooting lightbeams through a person's wrist and read the reflections. And it will need to do so millions of times within a few seconds. The amount of data generated by this method is massive, and that's where the tiny POET OEs come in (100/200G TxRx already done, 400G+ in prod by next year). With the OI, watch makers finally have a chance to capture enough sample from the blood to accurately measure blood content, especially glucose, while not having to change the current watch designs significantly.
The form factor is as small as your index finger's nail. That's why the OI is the holy-grail if things go right.