Potential counters that come to mind:
1) Make sure these relays never get built.
Some combination of always clearing threat out or zerging threatballs once these get to 50% to stop them. I'd kind of prefer a minimum strength for threat to start working on these, so I don't feel compelled to clear out tiny amounts of threat due to OCD, but that makes it easier to avoid the mechanic.
I think this encourages having one choke point with multiple ingresses; the threat will potentially spread out (dunno if threat is smart enough to clump onto one planet while waiting), and then you just sally forth and wipe out the threat as needed. Even if threat is smart enough to try to clump up, if you hold the center of an X map, they have to clump on separate planets if they come from different arms of the X. If you have multiple chokes, then you have to either split your sally-forth force or have it patrol between ingresses, which might be slow and/or require annoying micro.
Beachheading the adjacent planets would potentially help a lot, if you have VotM+Core Turrets. Makes great threat-bait at low risk. Without core turrets, you'd take ships from your main fleet instead, not take anything off your choke.
2) Kill the relays REALLY FAST after they show up
Gonna need a strike force ready when the relay is nearly charged. And firepower. Lots of firepower. Superweapons (ie artillery golems) would make this a lot easier, but having reasons to use superweapons for defense isn't bad. And then you're basically dedicating a strike force to defense, or having to turn around all/part of your main fleet whenever threat piles up on your borders. Then the strike force probably heads to whatever waves came in while the relay was up (hopefully, you know, none).
If the AI is smart enough to delay waves if a relay is almost done, or to launch waves early the instant a relay finishes, then this is much less viable; at that point you blow the relay and send the strike force to stop the waves, which might not be possible if 4 waves get launched at once (both AIs, two homeworlds).
3) Have multiple choke point planets, none of which are strong enough to deter threat.
This seems like one of the intended counters. You'll need more defenses to pull this off, though, either via unlocks or Core Turret controllers. See below for numbers.
4) Defense in depth
Reinforce planets far enough back that even if a relay gets built, your defenses can soak the waves on all those planets. IE, be playing low difficulty and have lots of Core Turret Controllers. This gets harder and harder the farther back those waves can penetrate, though.
5) mobile defenses
Just have half your fleet on defense all the time. That won't be a problem, right? This is how I usually counter WCA-GPs anyway; just pop them and move a fair chunk of fleet back to deal with the counterattack.
6) Play on maps that don't offer chokes
At which point you'd BETTER have core turrets and mobile defenses to survive; regular turrets will be spread thin and forts will have to guard irreplaceables. And don't turn on Shark.
Much as I love core turrets and VotM, it is a bit frustrating to have *the* major distributed defense mechanic be in an expansion rather than included as part of the base game. And even then, if you're on a big map and all the Core Turret Controllers happen to be really far away, you may be stuck without much distributed defense for quite a while. Miniforts only go so far, especially since there aren't higher marks available (hint, hint).
This all seems to have the issue by the wrong end, though.
If building one choke is really good, then maybe the AI just needs to be better at hitting that one choke REALLY HARD. Discounting core turrets, miniforts, and mobile defenses, if you go from two chokepoints to one, you've increased your defensive strength by 100%. Going from three to two boosts your defensive strength by 50%. I'm not sure how to interpret these numbers from the 5.036 patch notes:
* If 3, from (AIDifficulty*30) to (AIDifficulty*120).
* If 2, from (AIDifficulty*60) to (AIDifficulty*120).
* If 1, from (AIDifficulty*90) to (AIDifficulty*150)
If the base range of the wave is 3 mins, then dropping from 2 chokes to 1 makes the smallest waves 38% larger and less often, and the largest waves 20% larger and less often. That isn't anywhere near enough to keep up with doubling your effective firepower. If you throw a large enough wave at defenses, eventually the adjacent threat is going to go through. This might require annoyingly infrequent waves, though (you won't get feedback very often). Or the two AIs start syncing their waves if there's only one choke. Or they start deploying waves as cross-planet waves.
Conversely, splitting your forces across two chokes *halves* your effective defense strength. Any intervention intended to change that is going to need to be something like a factor of 2 or a total gamechanger to make it worth splitting firepower, unless you have core turrets and can increase firepower by splitting it up.
If the AIs always launched paired waves (like with two homeworlds), and the two waves launched always hit different planets if at all possible, that would make splitting up into two chokes way more worthwhile. Then the AI doesn't double up its forces unless you do, and having any core turrets at all probably makes it worth having two chokes as far as waves go.
None of those deal with threatballs intimidated by super-chokes, really. Nor do they really affect the power of using chokes vs CPAs, which are way more of a concern than waves. Maybe CPAs should take into account how many choke planets there are? Might be hard to make that un-cheesable, though.
Oh oh, here's a counter to entrenched planets: bored threatballs don't build warp relays, they build *beachheads*. Or they start doing cargo deliveries like astrotrains. Though an artillery golem sitting on your choke could probably one-shot beachheads easily, restoring your defenses.
Man, I'm glad I'm not the one balancing snake maps vs crosshatch maps. *salutes your dedication*
