I think they may have overstated the future threat posed by Indium Phosphide (which as yet has technical issues remaining to be solved)

Particularly, note this part:

<<It's an electro-absorption material, as opposed to an electro-optic material like the current crystals and polymers. Since lasers can also be made out of indium phosphide, there is potential that lasers and modulators could be integrated together into one lower-cost device with a smaller footprint. >>

Think about KVH's approach.

. No E-O chip
. Polymers in the core of the fiber
. A laser emits light through the "D" fiber, which connects directly to standard single mode fiber - light never leaves the fiber.

I can't see any reason that further integration of components can't be done - in fact, I'm sure they will. The polymers have lower crosstalk, meaning multiple devices can be placed closer together. I can envision future DWDM versions combining multiple modulators (and lasers) in a single package, which would also be lower cost and have a smaller footprint.

Also, economics comes into play. KVH's approach is extremely low cost, whereas InP will likely be high cost for quite some time (after they get the material working effectively).

In any event, this "threat" from InP devices probably remains at least a few years away.