A. Charge users from day 1. Your profit is easy to figure out, you should be profitable early.
B. Build a userbase, *then* figure out how to extract money from it.
A is a low risk lower potential reward, B is a high risk higher potential reward.
To get 'A' right, you have to line up product and price. That's not low-risk. Your argument that B is higher risk/reward than A is nonsensical; it depends entirely on the particulars you choose.
If A is failing, you can tell as you're going, and adjust. It obviously still requires skill, but it's just the time tested "make something, charge for it" model. As soon as you get users, by definition, you get revenue, and if you're competent, that means profit.
With B, you are deferring all or some of your income to some later stage. If it's not going to work out, you might not know that until the last minute when you're supporting a large userbase. You might never find a business model. OR you might hit gold.
Both models are just as valid as each other, it just depends on which you prefer, and which one is more appropriate to a particular set of circumstances.
