Hrm - I'm not aware of a specific policy against it, although we might ask someone to stop if the job were using up too many resources or taking down the server.
I know that some users have been using awstats, which is more processor / CPU intensive than analog, and in most cases, this doesn't seem to cause problems.
I'm not an expert on our stats setup, but I will look into it. Sorry for the short response before, but if you check the announcements page, you can probably imagine why I might be a bit busy / grumpy this morning.
I can't provide the best explanation of how the stats system works but I'll give it a shot.. hopefully someone else will jump in and correct me if I'm wrong. Our stats processing system doesn't run directly from cron; a lot of the processing is handled by backend machines, and we have our own scheduling system (which ultimately does involve cron and at) that handles scheduling of stuff. Because we have so many stats to process, we can't simply run everything at once. Our stats system is designed to compensate for this (to some extent) so that it doesn't need to run at exactly the same time every day.
I'm not aware of a general problem with this, but it does seem from the posts here that there is a problem affecting at least some customers. If you haven't already written support, please do so.