Here's one for anyone who uses Fangraphs' WAR as a source of anything: Corbin has a positive fWAR currently. His 110.1 innings of a 7.02 ERA while leading the league in hits and earned allowed is apparently just as much as contributor to winning baseball games this season as as been Alexis Diaz' 1.94 ERA in 41.2 relief innings.
Stuff like this is why I close my eyes and start shaking my head any time anyone talks about Fangraphs. Statistics have uses. They can also be overprocessed and twisted to the point that a McNugget looks like something straight from nature. The latter tells you that Patrick Corbin is above replacement level.
there's been lot's of criticism of using FIP in their WAR calculation. That's the root of the stat issue. Corbin's FIP is 4.96, so, with 110 innings, that 0.4 fWAR means he's been well below average but (barely) usable. At 110 innings, an average pitcher would have a fWAR of about 1.3-1.5. You are also right that WAR doesn't work well for relievers, too. WPA is probably a better stat for relievers.
You know who is dead last in fWAR among pitchers with > 100 innings. Our friend Josiah Gray, at -0.3. Reason? It's his homers. That pushes his FIP way above Corbin's (5.71). Corbin's fWAR is 4 lowest among regular starters. The guys who have a legit complaint are the 2 non-Nats, Jose Berrios and Marco Gonzales. Gonzales has an ERA under 4. Berrios has less of a beef (ERA 5.19) but he has a winning record, 8-4.
Finally, part of the reason Corbin's FIP is 2+ runs better than his ERA is his BABIP (.384) gets neutralized in the calculation. The team's pitchers' BABIP is .303, so Corbin probably doesn't deserve the break he gets on FIP. That's a long-winded way of saying Elvir is right.