http://play.prismata.net/?r=BW3ZB-qfvaK
So here's a game in which I do ignore Apollo and lose. Is it because the game went on so long (30 turns, although I'd lost by 20ish) that Apollo was very efficient damage?
I think you did two big mistakes this game.
The most important one was to buy multiple Xeno Guardians. Xeno is a good unit because he has vigilante. He can attack and block at the same time (and has 4 health). That's super good, and it's often a reason to get him as fast as possible, but you almost never want to buy multiple copies of him. Each one after the first one is just an overly expensive Tarsier. Think about it, aside from the first one, what does he do to make him more valuable than a Tarsier? If you block with more than one, you lose them, which you don't want to happen anyway. After the first one, buying multiple copies is actually even worse than buying Gauss Canons, as they cost 9$ and are basically equally strong.
The second mistake was not to get Apollo. Man, I never said not to get Apollo. What I said was, Apollo is not a reason for you to get Gauss Canons, because getting Gauss Canons doesn't protect you from him. Get Tarsiers over Gauss Canons, but then buy him anyway (if the game goes that long). Your opponent did have good targets, he had two Steelsplitters, which cost 7$.
Of course, in this game you do have to find a secondary damage source eventually, simply because there are just 10 Tarsiers in the supply. What I would have done is, get one Xeno ASAP and then look for the most efficient source of damage. Tarsier is one damage for 5$, that's the best. Get all 10 of them before you buy any other damage units. After that, the second best is one damage for $7 from either Steelsplitter or Gauss Canon, so start getting those. And add Apollo at some point during that.
By the way, if anyone disagrees with anything I say, please comment on it.
Your cost conversion is confusing me... You are converting every special resource into one coin. I think that is very misleading. One coin is not equal to any other resource. Heck, resources are very much not equal to each other. One spark, one green, one blue or one red are wildly different in "cost".
Anyway, Apollo is quite possibly the unit I understand the least (I think I've lost most games that involved him), so disclaimer.
I am still unsure whether tarsiers are a good target for Apollo, because they are cheap, and they make your opponent fear breaches. You can snipe your opponent tarsiers, sure, but he can flood you with them faster than you can take them out. And taking them all out actually removes a weakness from his strategy. I guess it depends on how long you think the game is going to be, and what your short term breaching perspectives look like.
That's a problem with all heavy-blue strategies though: if your opponent sees that you are going that way and goes heavy red, you are already heavily disadvantaged anyway. I think.
I disagree with your opinion that gauss cannons don't counter Apollo. The objective of going heavy green is to reduce possible Apollo targets to drones or conduits, while you keep hammering him with reliable attack. Even if he breaches, you don't really care that much, because green units have a lot of health (just don't let him destroy all your drones). Whether this strategy works depends on what else is in the board, of course. For example here, I think Xenoguardians are good. While they are expensive, costing 7 coins and extra blue (!), they are still immune to Apollo, and they defend you. It's hard to find something to defend yourself against Apollo, since they chew through most defenses (walls, rhinos, what have you) quite quickly.
You are right that gauss cannons are slow though, so if that's the only thing you can buy with green, it's probably better to go for Apollo yourself.
About Xenoguardians: you want to compare them to steelsplitters, not to tarsiers. If you want a tarsier, buy a tarsier. If you want a steelsplitter, see if you can buy a xenoguardian instead.
I don't know if it is that clear cut here whether you want to go Apollo or not. The Apollo player can use his extra blue for grenade mech and pixies (and xenoguardians). The green player can get xenoguardians to defend himself and keep the attack going... well that may be too slow now that I think about it. Meh. Ok yeah maybe you want to go Apollo here anyway.
One last thing, I notice that your opponent doesn't rush his Apollo, but instead builds up some Apollo-resistant attack first, which obviously delays his Apollo by 5 turns. It works out for him because qmech doesn't build any good Apollo target (walls, steelsplitters) until later. That's actually quite well done, I have to try to remember to do that.