I think the Stack Exchange system actually provides a bit of guidance in the matter. Consider the description of the no improvement whatsoever reason for rejecting an edit:
This edit does not make the post even a little bit easier to read, easier to find, more accurate or more accessible. Changes are either completely superfluous or actively harm readability.
I would argue that only changing from one style of spelling to another, without any other improvements, falls into that category. While the edit might not make the post harder to read, in the general case such edits fairly clearly don't make them easier to read either. They might for some subset of people visiting the site, but far from necessarily all people visiting the site.
Now, things get a bit more difficult if such edits are made at the same time as fixing other, more immediate problems with the post. At that point, it pretty much comes down to a sort of judgment call; is the post, on the whole, better with the edits than without them?
Perhaps the most correct way to handle such a situation is to pick the improve edit option, and change the spelling back to the original poster's style while keeping the improvements, but if the post is substantial it might be that the person reviewing the edit might not have the time to do so. This would be the correct course of action if the edits, besides the particular flavor of English used, are substantial enough to stand on their own. If I am not completely mistaken, this will award the +2 reputation to the editor for the edit, it will show as accepted and be stored as a revision, and a separate revision will be created with the reviewer's edits.
Alternatively, if the edit is mostly such style changes and only one or a few things that actually improve the post, pick reject and edit and carry the good changes forward. That way the edit will be rejected, and the editor can become aware of the fact that edits that consist primarily of such style changes are undesired.
Either way, always keep in mind that we all start out as newcomers to the network, and try to explain why you take the action you do in the review, so that the person on the other end of the keyboard can learn from the experience.