For someone working in the are it should be fast. It's too far away from my are to skim and find the errors.

Anyway, the default for a preprint (or paper) in the trash bin. Unless it's recommended by someone I thrust or it's published in a serious journal, I'd just ignore it.

For a post in HN, I can take a 10 minute look and try to find something interesting to comment about it good or bad. But not more than 10 minutes unless there is good reason.