Agreed that this could be improved (but I agree it's a P3).  Cc'ing a few folks that might be interested in grabbing it if they have a few spare cycles.

Actually, I didn't explicitly set the priority when filing. To get an idea of the impact, see b/118448076. Basically, users have no way to recover, unless they know about the chrome://extensions page, and they will be stuck with no way to resolve and start working offline.

Would it be difficult to trigger a Reload when the user click "enable" ?

The implementation is not terribly complex (though it might require coordination between the webstore and the platform, depending on how we approach it), but it is still a question of prioritization.  When an extension crashes, we will show a "balloon" notification that the user can click to re-enable it.  We'll also re-launch the crashed extension when the user restarts Chrome, and the user can also re-launch it from the chrome://extensions page.

So for this to happen where the user is stuck (and enabling from the webstore would help), the following needs to happen:
- The extension crashes
- The user does not re-enable the extension by clicking the notification when it crashes
- The user does not know to go to the chrome://extensions page and re-enable it
- The user *does* know to go to the webstore to find the (unlisted) extension

That seems like it's reasonably edge-casey to me (and is also solved by relaunching Chrome).  Do you have data on how frequently users are in this situation and unable to recover?

(To be clear, I'm certainly not saying it's not a worthwhile improvement, - I think it is! - just trying to get a clear understanding of where this might fit in with everything else we're working on :))

Thanks for walking me through the prioritization reasons.

The last point "The user *does* know to go to the webstore to find the (unlisted) extension", is actually more subtle. Since inline-installation has been removed, apps are now forced to redirect the user to the webstore listing to get the user to install or re-enable the extension. So really this is quite a common flow for us cause we prompt the user to re-enable the extension if we can't detect it.

That's a fair point, and I agree would make this more common.  I'm still not sure it's frequent enough to be urgent, though.  That said, if you get a lot of reports about this, or gather any data on it, please do let us know (we can always reprioritize if this turns out to be more common than I suspect :)).

Will let you know when we get reports. Thanks.