I'm reminded forcefully of the advantage IS gives whenever I've been able to place my non-IS binoculars on a rail or other location and get them totally stable, for the improvement in image quality has always been painfully noticeable. You can quite simply discern more detail - if I'm looking at something I cannot quite figure out, such as whether the peregrine partly blocked by a piece of masonry is still there, I always try to rest my binoculars against something and get them as stable as I can. I suppose I should note that IS is still not as rock solid as a tripod, but it's close - very close indeed I'd say if looking at the 10x42 IS-L.
Another great advantage that stability gives you is in ease of view, including in terms of eye placement, and the ability to observe with far less fatigue. The more difficult the target the more important all these qualities become. If you are one of those who prize the so-called "quiet view", IS delivers the near ultimate in "quiet" - so still it's like everything is holding its breath.
The thing about IS, though, is that in working so well, it shows up poor intrinsic image quality - its stability allows the eye to scrutinise the image it delivers more closely. Softness in the image, CA etc... are all more apparent, and as a consequence more annoying, than they might be otherwise. Because of the stability IS gives, intrinsic image quality needs to be high. Given that it's easier to make a good porro (which all Canons are) than an equivalent roof, Canon ought to up their game in that regard.
IS is an amazing function, with real utility in the field - there's no doubt about it. It's just a shame that the binoculars that have it are less than enjoyable to actually use, and there are questions over weatherproofing for all but the 10x42 IS-L. And I suppose for most birding one can do without it - by experienced observation of field marks the great majority of birds can be identified without needing IS. I'd like to know how much more effective users with long experience of IS think it makes them as birders - do they get say 20% more IDs than they would with "muggle binoculars", for instance? Does the IS device give effectively 20% better reach?
Now I think a stabilized scope of some sort, weatherproof, that doesn't require a tripod - maybe just one of those bendy-legged supports for GoPros and such - that could be interesting...