Well using cheap math you're actually fairly spot on with your %, and you will come to realize that its balanced to an "extent" given other parameters we can loosely control for.
- In any given game your win rate is 50%. ANy game is winnable at first.
- Sated Devour is a late game item thus the item itself has a success rate of probably 30%. This item alone you will probabilistically lose more games than not.
- Take you losing rate 50% * 70% of your loss probability with sated and minus 1= There is your probable success rate.
Here is why. Its a late game item that really can't tip the scales in your favor unless the match is fairly even or you are ahead. There is a 50/50 chance that the game relatively even to the point sated could turn the battle. There is a 25% chance that the enemy team is either way ahead or way behind. Sated really only effects that 25% you are already ahead. The 50/50 situation is such a large variable one team fight can make sated useless or not. If you are behind chances are you aren't getting stacks.
The issue becomes you have to make this choice early in the game where chances are you're in a 50/50 spot that can change at any moment. This is why high level pro play doesn't really value it because it lacks consistency. I would rather have an item that can capitalize of the 50/50 and push that advantage towards me. Your jg could get the kills in a team fight or they could be the first to go.
Chances are you witnessing your team being behind which makes their sated that much more salient. Next time see if you can spot this pattern in both directions. Especially competitive games that are dead even you will see how sated's vale shifts.