Mackwheaton, I fully agree with you that POET's "approach is to build from an elemental level (detectors, vcsel, to basic integrated devices and so on) and increase in complexity as each level of commercial production is attained".

But still J. Rodrigues said, that they expect that the problems with the FDA can be resolved. Bringing this topic up and being optimistic about it, gives a hint, that they won't abandon this product. One way could be, as you mention, that they could partner with someone.

So it would be very interesting to know more about their application at FDA. As this article "Iris as a reflector for differential absorption low-coherence interferometry to measure glucose level in the anterior chamber" shows they are already working on it at least since 2010. In the article it is written under "Materials and Methods": "Two different low coherence light sources were coupled by a wavelength division multiplexer (WDM) at the interferometer entrance. For this work we used two fiber pigtailed SLDs (Denselight Inc., Singapore) with center wavelengths at 1310 nm (FWHM bandwidth Δλ = 50 nm) and at 1625 nm (FWHM bandwidth Δλ = 50 nm), respectively." (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4570116/)