The plan as has dribbled and drabbled out from Musk's various statements is that the fairings are large and light, thus reentry heating is not too bad on them. They do not need to do a re-entry burn as the first stage seems to require, in order to slow down enough that hitting the atmosphere does not melt the entire structure.
Then once in thicker air, a parachute/parafoil is to be deployed to try and guide the fairing to a specific recovery area.
In the post Falcon Heavy video, Musk said that they are having issues with tangling the parachute lines.
Once down low and approaching the ocean, currently the fairings have been hitting the water. There is a ship in California with large arms attached that is postulated to somehow try and catch the fairing before it hits the water. Whether this is via a net or somesuch is still open to debate and further evidence.
All that said to explain the current approach.
Looking at the timeline for a Falcon 9 mission (Used Zuma since it came up first) we see that MECO (Main Engine Cut Off) and when the first stage starts working on getting back, happens at 2:20 in the flight.
Fairing separation happens at 3:08 in the flight. That is 48 seconds, which does not seem like much time, but in that interim, at 2:33 into the flight the boostback burn has already started.
In those 48 seconds the second stage has really kicked the horses and accelerated the hugely lighter mass of just the second stage + payload.
No matter how you delayed the boostback burn, the fairings will have a lot more velocity, be a lot higher, and quite far away, such that the first stage could not realistically catch up or loiter to collect the fairings.
Then ignore the issue of how you mate two very large fluffy objects to a first stage, all while falling deeper into the atmosphere further every moment.
This seems unlikely to impossible.
(I did look at some other flight profiles and MECO/Boostback/Fairing separation do differ slightly, but the example seems sufficient to make the point).
