First of all, you'd have to blame the poster of the specific deck, not the entire thread. This is a community, not a Wiki, and will have ideas that are flawed or experimental. Sometimes the poster is aware enough to know this, sometimes they're trying to get e-cred. Go to the Wiki if you want a shorter, more reviewed list.
More EA's than QT's? You only need one EA to enchant her entire stack of QT's. In total you only need to draw 2 the entire game: the second one for her Water towers. I've ran an unupped Immo counter against her on my alt for months, and sometimes you do lose because EA's cost 2
and she'll put up a Cloak before that. Despite this, it has a 70% winrate.
Now for some math: if we assume an upgraded card is worth 1100 (average is a bit over that) electrum and you win a card 30% of the time on average from FG's (which is undershooting it by 17%), you get 330 per win against an FG. Challenging an FG costs 30 electrum. This means a deck only needs a winrate of 9% to almost break even (99%). Do you see where this is going?
Now before we talk about RNG: if you play any decent counter a relevant amount of times, your gain will catch up. Even if your luck was the worst in the world, you'd still be close to evening out once you have enough statistical data. Here's something you can do to see how good a counter -actually- is instead of instantly raging after one match in a game that's controlled by RNG: load the deck in Trainer (elementsthegame.com/trainer), load the AI deck code for that FG and test away.
After a few more weeks/months on your real account, you'll also find most counters to produce a very nice amount of electrum per deck. Some of them are at 90% winrate. If a deck had that against all FG's, its FGei (basically a measurement of electrum/h for FG's) would be through the roof and it'd be the most profitable deck ever. You have the luxury of employing this every time you spin a specific FG.