These appeared within the last couple of weeks as
the new interchange at GA 316 and Harbins Road replaced the former traffic signal there. Apologies for the poor quality, but we were facing directly into the sun and it took every bit of photo-editing power my Galaxy S10 could muster to make them this clear. It also dramatizes the fact that there are visibility issues in addition to the comprehension issues I'm about to discuss.
This sign is at the beginning of an auxiliary lane added to develop the capacity of the two-lane, two-destination offramp ahead. There's an MUTCD-compliant way to sign this condition-- as it happens, it was used
a few miles up the road a couple of years ago. Nobody really likes it, though, and GDOT usually does something else, such as
this or even
this.
Here, though, GDOT decided to go with an MUTCD-defiant unisign, a concept that, as I understand it, proved inferior to APL's in whatever testing was done, which is why APLs are in the MUTCD and unisigns like this aren't. The divider line, rather than being placed directly above the center arrow to indicate an option lane, is located to the left of said arrow, indicating, incorrectly, that two lanes go to the offramp and only one continues on the mainline. Really, it would've been more accurate to use a conventional (MUTCD-defiant) sign
like this with an arrowless pullthrough. I guess that they did it this way because they really, really wanted to inform drivers that the offramp splits later on, causing the Sugarloaf Parkway text to displace the dividing line from its proper location. Ironically and infuriatingly, the Sugarloaf Parkway branch of the offramp is closed, with Sugarloaf traffic using a temporary offramp in the vicinity of the bridge we see in the distance.

Next comes an APL at the exit divergence. Again, the Sugarloaf Parkway text displaces the divider line to the left of its correct location. Note to designers: The dividing line on an APL
ALWAYS goes in the crotch of the split arrow. Here, IMO, they would've done better to omit it altogether. Or, better, they could've gone with
Sugarloaf Parkway
Harbins Road
and not had a problem.
Also, as I mentioned before, the Sugarloaf Parkway branch of the offramp isn't open, and drivers bound there need to stay on the mainline until the vicinity of the bridge. They should've blacked out the curved part of the split arrow, not the straight part. WTF were they thinking?
But, wait: there's more: There's no reason to encourage drivers to use both lanes of the ramp because the ramp enters a tortuous one-lane detour shortly after the not-yet-operational split. Wow.

One more thing, and this is design rather than signage: They really shouldn't have combined Sugarloaf and Harbins onto a single offramp. As you can see, it needlessly complicates signage and operations. Besides, they're separate exits in the other direction anyway.
One more one more thing: Sugarloaf Parkway, a Gwinnett County project completed almost ten years ago,
also uses unisigns, and for no apparent reason. Perhaps they had a hand in this mess.