I probably should've made this clearer, but the 50/50s we are referring to are very crucial plays and really only occur late in battles when it's like 2-2 and that is where the coinflip comes into play. The thing is, those coin flips don't only apply to genesect, they can happen with any other mons as well.
exactly what i'm trying to say, true 50/50's rarely happen outside of the end of the match, due to the odd circumstances, which i'll just touch upon later in the post
Lets say we are talking bug, and a magnezone kills your lead shuckle with a flash cannon. You then proceed to bring in your genesect, lets say it is indeed scarfed, but your opponent doesn't know that, he could be thinking it's scarfed or life orb or specs. Lets say the magnezone is specs locked and does not want to stay in and get popped with a flamethrower. Your opponent is most likely going to switch into his/her zapdos to sponge the flamethrower and threaten back with a heat wave or a discharge. Now the question is, do you flamethrower and let zap get in for free and fire off a heat wave against mono bug? Or, do you bop the zapdos with an ice beam and cripple it for the rest of the battle. There are TWO scenarios to that situation, so its a 50/50 right? No, that situation by know means is a 50/50. That is simply you making a choice on what you think your opponent is MOST LIKELY going to do. If you get it right you are rewarded with a near dead zapdos, you ice beam and he stays in, you catch a powerful flash cannon to the mouth and gene is taking close to half. Now that play is not going to cost anyone a game which is where I think the mix up is. Those 50/50s are late game 2-2 occurrences which happen all the time and should not be applied to gene.
this is a good time to point out what i said previously: a 50/50 is a situation in which there are, usually, 2 plays for each player.
All 4 plays (2 for each player)
have the exact same risk and the exact same reward as the others. Let's take your example. Lemme try and list the risk and reward for each play without using a table because lol effort.
Genesect user
Flamethrower
Risk
Letting Zapdos come in for free and getting a hit off on your team. The degree of this risk varies depending on your team: e.g. if you have Armaldo/Volcarona, the reward is very little, as the attack will not do much damage at all. If not, there is a larger risk in letting Zapdos come in for free, making this play slightly less safe than before.
Reward
KO'ing Magnezone (or if you can't KO it, heavily damaging it). Degree of the reward depends on both of your teams; if nothing on your team can stomach a hit from Magnezone, then this is a very high reward for you. However, if Mag is outsped and KO'ed by nearly all members of your team, or is hard walled by something (Bulky Rona), this is a small reward and the value of using Flamethrower goes down. Now, if Mag is the opponents only answer to a mon on your team, and he needs to keep it healthy for this reason, there is an immense reward; opening up a position for that particular mon and helping it possibly win the match entirely.
Ice Beam
Risk
Letting Mag get off a large hit on you and weakening Genesect to a very low number, and subsequently getting of a hit on the incoming mon. If Genesect is your way of winning this match (breaking through opposing team or cleaning it lategame), then this is a large risk (don't forget to take into account if you need Gene healthy or not; if SR/Spikes won't be an issue in the match and you won't be using it to stomach any hits [priority] then Genesect doesn't really need to have its health for any reason in particular and the Risk of this play drops dramatically). However, if Gene isn't really that needed in the match, and you have a clear cut (or close to this) way of winning even if Gene is weakened, then there is low risk. Also think about the turn after Gene gets Flash Cannoned; if something will die to the next Flash Cannon (Gene can't stay in; is locked into Beam), then the Risk of such a play is generally very high. Even if you have a useless mon, it's nice to keep that for later in the match.
Reward
Crippling Zapdos and subsequently KO'ing it or getting a big hit off on another mon. Depends on teams; how important Zap is at beating this Bug team, how many Ice Beam switch ins the opponent has, etc (shortening this due to lack of time)
Ok i dont have enough time to really get into this (i would do mag plays too), but you get the idea. if you notice, i gave multiple scenarios where the risk/reward increased or decreased dramatically. my point is that all of this heavily depends on the team, since with 6 different mons you get a lot of different plays and the risk/reward changes a lot. this is exactly why 50/50s rarely happen outside of the endgame; there are just so many factors to take into account in early game and midgame that it really just can't happen. hopefully the example above clarified how risk/reward changes depending on the specific battle, and when looking back on 50/50s (where the risk/reward is literally "lose game, win game"), it clarifies how unhealthy they are, since this cannot be used in those scenarios and these "predictions" are reduced to a mere coinflip.
As far as predictions dictating the course of a match, Yes they most certainly do. I probably did exagerate in saying they make or break a battle because those are the late game 50/50s, but you did prove my point for me in saying that they are determining factors in playing around win cons and and all the other stuff. That is more or less what I was trying to get at. The situation above is a perfectly healthy situation and should not be put against genesect.
yea predictions definitely dictate the course of the match, but i would say not as much as long term thinking and overall game plans. however, that isn't what we're here to discuss. if the above scenario is all that genesect forces, then it is by no means broken by "creating too many 50/50s". that scenario can be caused by several pokemon that simply have decent coverage.
i'm not gonna give my opinions on whether or not gene does force these unhealthy 50/50s, due to lack of experience battling genesect, but there needs to be more than this example, or the one scp posted above (char y vs gene), as those are not 50/50s at all; guesswork can be completely avoided by simply thinking out risk reward (which i might add scp did very well there; risk of switching out is facing heracross, who could possibly get multiple kills. risk of staying in is a ko on each side. reward of switching out is getting a kill, reward of staying in is getting a kill. since reward is the same, there is less risk with staying in, so that is clearly the better play)
replays will be immensely helpful in arguing the point that gene forces 50/50s, since its difficult to theorymon this and it is much better shown with a replay. just pointing out again that i am not going to take any side of gene ban, due to lack of experience, but i most certainly will step in when people get things that are healthy and unhealthy mixed up, as that could be detrimental when creating a balanced meta