So from what I can see you have 2 ways of killing a mecha'thun (that gives you a win obviously). You can either make it 1 mana through Jepetto, which is unreliable and obviously not the main plan of the deck and then you have what I would actually hope is your main plan of copying mecha'thun with floop which is obviously much easier to kill.
This second gameplan is perfectly reasonable on a vaccum, but it's kinda not so good because you only have wrath, starfall and bes to kill your mecha'thun, and the first 2 cards are pretty prime removal you have for opponent's minions so you might not have them to kill your floop. Other than that I don't get why aren't you running Kun in this list, it's literally a free 7/7 that heals you for 10 hp, so it means you can drop an extra card from your hand for cost 0 which gives you a 7/7 and 10 health, pretty solid for literally any druid deck that plays the quest (which are like 99% of them). Druid got rekt with naturalize being HoF'ed it was such a nice utility card but now you have to put on a bunch of extra work to kill the floop and you can't kill the actual mecha'thun consistenly.
Having said that, on a realistic meta environment, I don't think this deck would really work. As druid you have a solid amount of draw, pretty much any druid with quest will fatigue on most games but you still have to play everything in your hand and then spend a whole turn dropping mecha'thun and finding the means to kill it(has to be played as the last minion or you'll ruin your floop) and save both floop and one spell that can kill it for the entire game, that just seems way too slow for the current meta. With decks like evolve shaman (quest and questless) and all the multi revival meta decks rn (priest and multi n'zoth stuff like rogue and warlock) you simply just succumb to all that pressure before being able to drop your entire hand and then still needing an extra turn to play the mecha'thun.