I come from a background in copy editing, so (not surprisingly) I think that a post with the word clothes in it is significantly better than a post with the word clohes in it. And as several experienced and well-informed editors point out in comments above, once you're in for the typo, you might as well clean up some of the other shortcomings of the post as well.
I didn't start editing other people's posts on this site until I had reached the rep level at which my changes skipped the review queue and took immediate effect, so I missed out on useful feedback I would have gotten from other users' vetting my suggested changes. But even so, I can't recall having received more than a couple of negative reactions from posters about the changes I've made to their posts. (Those arose in instances where I thought I understood what the poster was trying to say on an ambiguously expressed point and tried to make that meaning explicit, but I got the intended meaning wrong. In those instances I would have done better to query the poster about the intended meaning than to introduce changes based on guesswork.)
Because spelling, formatting, and grammatical corrections make the poster's question or answer look and sound better, a poster usually has no reason to react badly to such improvements. After all, no one wants to have a serious question or answer dismissed because it contains typos or inadequate formatting. On the other hand, altering the poster's voice or changing the poster's argument to reach a different conclusion—even when you think the change is for the better—is inappropriate at this site, just as it is in the world of professional publishing.
The only time I feel uneasy about correcting typos and doing other cleanup chores on a post is when the post has no substantive merit and/or has already drawn close votes. Occasionally, a post's faulty spelling and bad formatting obscure an interesting question or answer, and clearing away the rubbish makes the hidden issue or insight visible. But far more often, the post's shortcomings in spelling and presentation are entirely consistent with its low overall quality. In those cases, it doesn't make sense to try to improve the post on a formal level, since its substance will remain weak and close-worthy despite the upgrades.
You never know how a particular poster will react to editorial changes or suggestions until you interact with that person. But if your suggestions are usually sound (which you can tell by the proportion of them that are accepted in the review queue) and if the poster isn't unusually prickly about changes, the process is painless for you and beneficial for the poster and for the site.