I heard that March is the best month to launch, and April through December are very bad. I guess you will have to wait until March. :)

Actually, everybody always says a.s.a.p. Most people think they're competing against "people who will rip me off as soon as I release my product" which leads to conclusions of "if I release on day X, I will be ripped off Y number of days later, so I should wait until Z, the perfect time, to release." In fact, there's a higher chance that there are startups and current companies doing something similar REGARDLESS of YOU.

A lot of the reason existing companies have not caught on to idea X is because the whole reason you have idea Y is because the current companies just came out with idea X and there's simply no way to do Y before X. So, they may have already had your idea Y in mind, or started working on it, but had to release X first. So, when you got your idea, so did employees of all of these other companies, as well as budding entrepreneurs. So, you should release and at least get credit for being first. The whole point of being a startup is you can create whatever there's a huge chance big companies are working on as well, but much faster because you're just implementing basic features and avoiding bureacracy, as well as have no responsibilities. So, if you're trying to do a startup PG-style, I would try to create it and release it as soon as it works, with the bare features in it, and get publicity about it.

I'm sure there are other mantras, like targeting a very small niche and improving the product, and only publicizing it when its ready (but you have tons of feedback from your small niche). Or, maybe getting a lot of venture capital straight off and partnering with big shots after you're much older and have many contacts. But I think if you're doing it PG-style (and I don't speak for him), you should release it as soon as it works. More or less.

Depends on the site. A site offering customized Christmas cards won't do well immediately if you launch in January; a site offering something useful for preparing for college won't do well immediately if you launch in November.

Launch the site as soon as it's ready to be launched. You'll drive yourself crazy trying to micro-optimize and asking yourself if you'll get more traffic if you launch on Tuesday or Saturday.

I know I know -- as soon as it's ready to launch. But seriously, is launching a site going into the summer season a good idea? You're just not going to reach as many people as you would in the fall or right after the holidays.