Well, no one ever said that a manager doesn't have any impact. The impact of in game decisions is just not as significant as you think it is. I would have gone with Kimbrel once Barnes walked the first two batters because it was the % move, IMO. And even though the % difference in win expectancy is not large at all, that doesn't mean you shouldn't go with the % move.
That said, keep in mind that every move has a not only a positive impact, but also a negative impact in some other way. For instance, if Kimbrel is used in the 8th and can't pitch in the 9th, then there is a negative impact in the 9th. This makes the overall impact of a manager's move even smaller. Therefore, it really is not correct to say that Farrell cost the team that win. It takes way more than one 'bad' decision to cost a win.
It goes against the human way of thinking, but it's all about win expectancy and the inability to know what would have happened in the game had the other decision been made.
If you're interested, here is a very good read on manager's decision making impact with some actual numbers:
https://www.beyondtheboxscore.com/2015/10/21/9574081/the-impact-of-managerial-mistakes