Jeff Clement: .341/.442/.591
Gary Harris: .236/.263/.345
Adam Jones: .192/.250/.385
Clint Nageotte: 17 IP, 1.59 ERA, 9.00 K/BB
Chris Buglovsky: 12 IP, 12.75 ERA, 1.00 K/BB
Mike Flannery: 12.2 IP, 2.84 ERA, 3.00 K/BB
What's it all mean? Essentially, nothing. Against standard competition, a .297 hitter (Jones) stands a 6% chance of hitting .192 over a given 52 AB span of time. For the sake of comparison, this past season Richie Sexson went deep in 6% of his plate appearances, and I didn't hear anyone freaking out whenever he hit a homer. Beyond that, though, the AFL can't even be considered a normal league, because so many of the players are sent there for a particular purpose (throwing a new pitch, learning to hit a changeup, or - in Jones' case - getting better in center field). The combination of small data samples and unusual styles of performance make the resulting numbers pretty worthless.
Obviously, I'd prefer to have all six Mariner players performing incredibly well, but even if that were the case, it wouldn't give us any new information about the players in question. Some guy named Brian Wolfe, who has a 5.35 ERA over his last three minor league seasons, is currently posting a 13:1 K/BB in Arizona. The best ERA in the league belongs to Steven White, a 24 year old in the Yankees system fresh off a 6.44 ERA in AA. Last year, we saw fantastic performances turned in by RD Spiehs, Wes Wilkerson, and Jeff Housman, a trio who went on to allow 181 runs in 280 innings this past summer. By the same token, there are good pitchers who sucked, good hitters who sucked, and bad hitters who excelled. I don't need to go through a laundry list of examples.
Could Jones' struggles and Nageotte's success be indicators of future performance? It's possible. It's just not very likely.