This uses a flawed assumption. That any Pathfinder-style approach for 4E would use the GSL rather than the OGL.
No way around it, really. Pathfinder essential /is/ 3.5 - it's completely compatible, you can walk into Pathfinder with 3.5 characters, and you'll just suck a little because it's been power-creeping for a few years.
You can't do that with regards to 4e under the OGL.
You can't copyright game mechanics. Which means that all you actually need is a 4Eish game that doesn't do anything to copy trade dress and is written under and using the OGL.
That's harder than it sounds, since key things like the AEDU progression, stat blocks, and power blocks arguably fall under 'trade dress.' And, where Pathfinder could re-print core classes whole cloth, 4e classes you'd have to dodge around extensively. The result is unlikely to be compatible. A compatible result would likely be subject to a legal battle - and Hasbro needn't have a strong case to intimidate a small 3pp out of publishing, indeed, it's likely no one will even try.
It's much safer to do what 13A did, and create just another d20 games with just a few elements that resemble 4e, but 0 compatibility.
So, while there will presumably still be some good d20 games being made for the foreseeable future, and a few of them may use a 4e innovation here or there, there will never be a 4e clone to outsell 5e the way Pathfinder did Essentials.
5e's perfectly safe in that regard. The lack of any prospect of ongoing 4e support will help convince 4e fans (who, afterall, included D&Ds cohort of early adopters, as well as fans of the system, itself) to migrate to 5e - and give 4e hold-outs nothing much to hold out for (or keep talking about).
5e still has to watch its back around 3.5 fans, of course. It's best bet is to just hope that they see 5e as WotC 'taking their side in the edition war' and embrace it as a symbol of victory, even if they don't particularly care for it.