This seems to assume that the only reason to start a product company is because you want to gamble on the slim chance that you'll get rich. But what about starting a product company because there's a problem you believe really ought to be solved, and developing a commercial product is the best way to solve it? In that case, wouldn't it be prudent to avoid unnecessary risk?

You'll still most likely fail! This is a really important part of why the VC math works the way it does. You can de-risk a product company by not trying to shoot for the moon, but it's an incremental derisking.