So, I was taking a look at the home/road splits in Mariners' games this year, both for the M's hitters and for the opposition. The results surprised me a bit (stats through last night's game, thanks to baseball-reference.com; including tonight's game would only increase the differences):

Sea@Sea .196/.275/.288 (.563 OPS)
Sea@opp .256/.306/.417 (.723, difference .160)
opp@Sea .221/.282/.331 (.613)
opp@opp .267/.331/.457 (.788, difference .175)

So Seattle's opponents have, it seems, lost more OPS points to Safeco than our own hitters have (though, granted, the opponents had a few more to lose in the first place). Or, to put it another way, Safeco appears to have helped our pitchers this year about as much as it's hurt our hitters.

Drilling down on the elements of run production, I find:

Sea BABIP .242 here, .299 there
Opp BABIP .265 here, .300 there

I don't know any sources of HR/Fly splits, but looking at HR% defined as HR/(AB-K), I find:

Sea HR% .021 here, .041 there
Opp HR% .032 here, .054 there

Differences in K/W were less significant contributors, and showed no clear Seattle vs. opposition trend (Seattle hitters were a bit better on K/W at Safeco, opposition hitters better on K/W elsewhere).

So anyway: All of this appears to me consistent with the hypothesis that Safeco, for whatever reason (presumably weather-related), really has been that bad for offense this year-- that it's not just our hitters panicking over a few fly balls dying at the track, nor is it just a small-sample fluke. Everything it's doing to us, it's doing to the other guys too (and in more or less the same ways). If it were primarily a sample fluke, I'd expect wilder swings between the run-production elements on the offensive and defensive sides.

My current working hypothesis (somewhat to my surprise), then, is:

(1) Seattle's offense is not as bad as it has looked this year.

(2) Seattle's pitching-- and defense (note the 35-point opponent BABIP difference)-- are not as good as they have looked this year.

(3) Comparing the road stats (which I'm suspecting may reflect the team's true park-adjusted performance pretty well) with the AL average of .256/.322/.412 (.734) suggests to me that our hitting may actually be better than our pitching/defense this year (both a little worse than average).

(4) The team may have better reason than I had previously thought for not shaking up the position-playing side of the roster at the break.

(5) There may still, however, be good reason for adjusting the fences to produce a more neutral park; the avoidance of bad habits by the hitters might still help them to become high-quality performers both home and road. (I note, of course, that shorter fences would help HR/Fly but could hurt BABIP a little; but the net effect would surely be positive for offense.)

Comments welcome. I know this is counter-intuitive-- I'm seeing the same thing that you are, I'm not blind. But my hypothesis is that our eyes may be deceiving us. (If anyone has better split data than what I've looked at so far-- Infield Fly%, perhaps?-- perhaps that, or something else along those lines, would revise my conclusions a bit.)