From ourairspace.org*:
The FAA wanted to implement Continuous Descent Approach (CDA) at La Guardia as part of the Regional Airspace Redesign to reduce the noise impact when they terminalized a much larger area. In the Draft Environmental Impact Study (DEIS) CDA was included however CDA failed the safety field test in the summer of 2007 and was not included in the Final Impact Study (FEIS).
(...)
The FAA decided to rebrand CDA due to the safety and noise risks discovered overseas and in litigation so they changed the name of the procedure from “Continuous Descent Approach” to “Optimized Profile Descent” (OPD).
Note it says, "... was not included in the Final Impact Study," which to my understanding means the specifics did not make it to the final report (I tried to look for that report, others may have better luck).
They mention failing a safety test, and similar overseas issues. From this paper (atmseminar.org), the one thing that stands out is the need for 4D (time based) traffic management to make it work in high density areas.
This paper has shown that the conflict-free continuous descent approach flown in high density traffic can be achieved based on the 4-D trajectory concept. The proposed scheduling algorithm is proved to be effective in strategically solving the conflicts occurring in the inbound traffic in a full day simulation of terminal airspace operations.
Which means relying only on what each FMS wants is not enough. The ATM systems need to account for all the airborne FMS's, and based on that delays can be introduced for proper management. Not an easy feat.
For comparison, in Europe the CDA and CCD in high density traffic still does not have an IOC (Initial Operational Capability) date. One of the building blocks is expected in 2023.
* [A] volunteer organization of professionals who are committed to addressing and reducing aviation noise and emissions while supporting a safe robust aviation industry.