Fuji OIS Oddity?

I have exactly one Fuji X system lens with OIS, the 18-55mm. I like this compact zoom far more than I thought I would and use the crap out of it. I have a ton of hours of in-field use as well as more than a little bit of casual testing. There's one thing I just cannot get a handle on though — performance and predictability of the OIS.

Wide

Background Diversion

I've had stabilized lenses before, I've used other people's stabilized lenses before, I've even debugged other peoples issues related to their stabilized lenses. I've never really trusted the stabilization, trust it in the same way as I do other mechanical camera functions where I know exactly what's going to happen. I've never really tried to trust it, nor have I spent a lot of time figuring it out.

The couple of times I tried to nail down exactly what was going to happen under what circumstances I've got results that baffled and confused me. Not just on the Fuji but other brands as well. Example; A friend that does a lot of travel and nature photography was coming home with blurry images from made with his 70-200mm stabilized Nikon zoom. He swore it was some sort of focusing issue, or his lens was broken, or it needed focus fine tuning (which he'd already done a million times). I had another suspicion, the VR.

We spent an hour or two and proved beyond any doubt the poor results were entirely due to forgetting to turn the VR off when using a tripod. In fact the only time I've ever seen stabilization to reliably produce a particular set of results was the Nikon VR's ability to produce terrrible blurry images every single frame on a tripod with the VR turned on no matter what the shutter speed was. Too funny. Yes, I and my friend know you are not supposed to have it on when locked down on a tripod, or you are supposed to switch modes to the "other mode". Neither of us predicted the results of every single frame being rendered unusable if it was on.

Fast Forward To The 18-55mm XF OIS

I'm one of those photographers that really likes to know exactly what my camera/gear is going to do under any given set of circumstances. Sure I do hail-mary's when I've got no other choice but anything that produces variable results I file into a category of not really useful. Having owned the 18-55mm for more than a year I've tried to get a handle on exactly what OIS will and won't do. I can't seem to nail it down so I tend not to rely on it.

Wide

Example: On many many occasions I've tried to figure out where it breaks down, where it shines, and how it compares to my ability to hand-hold a camera without it. For the life of me I just cannot seem to figure it out. Let's take one particular set of tests at 55mm. I can hand hold 55mm all day at 1/125 without OIS and get the exact same results with OIS. I seem* to get better results most of the time at 1/60 with OIS on. Same for 1/30. Pushing it into the really slow zone I get bizarre results at 1/15th completely unpredictable, I can't even correlate it to my particular lack of stability shot to shot. Here's the surprise — at 1/8th I get better results???

The Oddity

Okay with all that water under the bridge I tend to leave OIS on most of the time when shooting between 1/30 and 1/125 because when I set out to test it doesn't seem to make a difference at higher speeds but increases the odds of a decent shot at shutter speeds I'd not normally consider reliable. Maybe.

The oddity is that when I make pictures under conditions that are no-brainers hand held, say 1/250th at 35mm or 55mm I occasionally see results that are bad. Bad as in blurry when they shouldn't be. I use the 18-55mm enough for me to be fairly sure it's not me at 1/250th and I see these enough to suspect that when I shoot at higher shutter speeds with OIS turned on if I'm extremely stable that OIS is actually causing the blurry results. I'm sure it's not focus and it happens enough that I'm 90% sure it's having the OIS on that's the problem. It doesn't happen enough where I can reproduce those results on-demand.

The first screenshot in the post is an example of what I see when I think OIS might be causing an issue. The second screenshot is more "how it should be" all things considered. Go ahead download them and look at it at 100%, or better yet don't bother. You'll probably have a difficult time figuring out what's going on. Especially give you can't move it around and figure out the lack of definition in the first screenshot has nothing do do with focus, etc. Take my word for it. They are very different and as far as one can figure it out have similar focus planes. Just stuck them in here to break of the text a bit, not as any sort of "proof".

Yep, simple solution if I am 90% sure. Just turn OIS off at higher shutter speeds. Too bad I don't remember to do this when in the middle of making pictures and need to go up a notch or two in shutter speed.

Here's my question; Anyone else out there patient enough or have enough experience to say definitively or semi-definitively if OIS is causing IQ issues at higher shutter speeds when hand held in circumstances where there should be no issues?

Curious?