I'm assuming you're talking about the very dangerous level 2 Embargo-Sova rush strategy that seems so common in 5s nowadays.

Scout manually and quickly. If you see a Sova coming, don't build a second cap, just use your first one; maybe scrap your cap factory for a bit of extra funds that you'll need.

If you're Advent, built a lot of Disciples ASAP. Build another frigate factory on your HW to produce faster. 20 to 30 Disciples, properly micromanaged, should be able to scare the Sova out of the gravity well quickly enough. If you're TEC, it depends on how much space there is between you and your enemy(ies). I'm not quite sure about this, but since the incoming Sova probably has fighters to kill constructors and LRM, and since labs/research costs a bunch, it might be good for the defender to build a ton of Cobalts, very similar to the Advent strategy but slightly less effective. However, depending on how much space you have, you might have enough freedom to research and build up enough LRMs to scare the Sova out quickly enough, which would be more effective later in the game after the rush is over. Vasari? 1-lab Assailants plus phase missiles should be good unless the enemy is like right next door.

It would help to use your cap (a carrier, right?) to defend as well, mainly for strike-craft support. With fighters of your own taking out enemy fighters, building LRF will be a more viable strategy.

If the enemy is good at MMing though, all of that sometimes won't be enough: especially if your frigates are LF, you might be having to continually ordering your frigates to move closer to the Embargo-ing Sova so they can fire. I was recently playing a game with someone and homeworlds were right next to each other. He quickly built around 15 Disciples and kept building them (up to 30) and had them attack my Sova (regular attack). However, the Disciples seemed to have a very hard time maneuvering around to fire: I was carefully micromanaging my Sova's movement around the planet, and the Disciples had a tendency to get in range, fire, slow down, and then scatter and try to get in range again. I think the auto-attacking AI has maneuvering problems especially when it's right next to a planet or if it's going through a pack of orbital structures. My Sova managed to stay alive and in the enemy homeworld well for 15! minutes before it was forced to retreat and repair. If I had been supporting it with LRM backup (and/or another Sova) the game would've been over for him.

Therefore: If you're using light frigates with short ranges, if the enemy is MMing their Sova around your HW you probably need to be continuously reissuing Move orders to past the enemy cap to maximize the effectiveness of your LF.

Long range frigates will go much slower (they go exactly the same speed as the Sova), so if you use them they may have to move across a large portion of the gravity well first before they get in range of an enemy Sova. That said, due to the slow turning speed of the Sova and due to the long range of the LRM and Assailant, once the Sova is in range, your LRF will probably easily be able to continuously fire until the Sova leaves the gravity well - and will also be doing more damage.

Obviously, you should be only targeting the level 3 Sova with L2 Embargo. Since training to level 3 costs a lot, it will take a good amount of time for your enemy to get enough income to research and build another Sova. I'm not sure on this but if you use these strategies you might be able to cripple the L2 Embargo Sova and put it on the retreat before the other cap arrives.

After the Embargo-Sova retreats, follow through! You probably have LF, which are fast and are sure to quickly get in range to attack the Sova (especially when the Sova is going straight towards its HW without trying to evade). Even if the other cap is bombing your HW, don't worry about it for now, your HW has a lot of health.

By the time the Embargo-ing Sova retreats, you should have a sizeable number of LF, probably enough to do 75-100 hull damage to the Sova a second, which is a good number. Follow it into the next gravity well and pursue it all the way to its homeworld: if you're good at MM, often it'll be possible to kill it before it can make it back in time. (either that, or a good Sova rusher will be forced to retreat before Embargo has done all that much economic damage to you)

If the enemy's tried a Sova Embargo rush and is attempting to build 2 caps like you said, they'll be quite short on funds and might not even be able to make effective repair platforms in time, especially if they're trying to build a supporting fleet of LRM. If you can take out the capital ship, you'll have a very significant advantage even when the Embargo economy effects are accounted for.

If you are in 2v1 with 3-4 carrier caps on your HW within 5 min of the start of the game, it's only a matter of time before you die, feed or no feed.  The exception is if the two people suck very badly, but even then that strategy is hard to counter.

How did that work? Unless it was an Orkulus, couldn't the attackers just fight/bomb from outside the range of the SB? And even if it was an Orkulus, it's not that effective against large single targets like capital ships.

Ok thank you for confirming this. I did not see a way out of it at the time and wanted o make sure that I respond correctly when it happens to me. 

This one will depend on your may layout.  Generally speaking if the map layout has fewer lanes then offensive culture will be much stronger and your opportunity to counter-attack will be quite limited.  This would be a great killing strategy in such a circumstance.  If the map is more porous, culture will be dispersed and you'll need a ridiculous number of culture generators (perhaps as many as 8) to overcome a couple capital ships on defensive duty and a single defending generator.  Furthermore, the defender will have more counter-attack options and a better opportunity to grab a few colonies of their own.

This works both ways; the defender can knock out the attacking structures as well.  A good defender isn't going to sit on his homeworld twiddling his thumbs... and if he does do this, you really don't need to kill him at all.  If he's good, he's going to be out there hitting undefended flanks and keeping both players on their toes, preventing one from economizing or turning on another front.

Maybe if the devs were willing to scale down the cost of culture SB upgrades to a cost comparable to what they're worth...?

Culture bombing with SBs can be a huge drain with little return on small games....but on larger games, I've found culture SBs to be very very effective, especially when trying to deal with an entrenched player....

I'd say that the opportunity cost of three SB upgrades is more concerning than the actual upgrade cost...but, in large games where players specialize in roles, I think it is a pretty balanced mechanic overall and a very nice Advent card to play....