Baseball Reference also has statistics for clutch, and, like Fangraphs and unlike everyone on this forum, they have a definition for clutch. Around here it always boiled down to "clutch exists but I don't know what it is."
And my argument ABSOUTELY DOES NOT completely come down to statistics. You proposed that dWAR was useless because you did not agree with one conclusion. And no proponent of WAR, bWAR or FWAR, has ever said the number is definitive.
And WAR is solely used for historical data. No one has ever used to to predict the outcome of a single play. And why would you? It doesn't look at a single aspect of the game, so using it to predict one would be silly.
OPS is a stat that is completely unquantifiable, too. If I asked you what the difference was between an .850 OPS hitter and a .900 OPS hitter, beyond saying the latter hitter is better, you couldn't do it. But when you break down the components of OPS - OBP and SLG - you can get an idea of which hitter is better. The normalizing OPS+ is an even more confusing stat, since none of us can even calculate it, and certainly useless for a single play. Does that make it unquantifiable? Does it give no relevant information?
But these stats are all usefull for comparing players we don't get to watch to the ones we do. And just because one stat says a player you think is doing well is doing average doesn't change how good he is or invalidate the stat. This isn't about "eye test" vs raw data, especially since most of these stats do incorporate eye test data along with defined standards (that casual fans do not use).
These numbers are best used in comparing players. Look at Mookie Betts. In his rookie year, he posted an OPS of .812 and an OPS+ of 126. In his second year, he posted an OPS of .820 and an OPS+ of 117. His OPS improved but his OPS+ decreased. What does that tell you? Did Mookie get better or worse from year one to year two?
The answer of course id that while Mookie did improve a little, but the league around him improved more.