>A: Yes, you can make money; however, you have to get in as soon as >possible, what are the chances of that?
I'm not sure about this. Your scenario assumes everyone complies with the scheme. But in point of fact, when anyone receives such a chain letter, he doesn't have hundreds or thousands of companions, only those whose predecessors chose to continue the chain. The majority of the recipients did not participate. So any recipient is actually getting in near the effective top. In other words, since so few people choose to participate, the chain can go on for a very long time indeed without exhausting all the people in the world.
On the other hand, with a small percentage of the recipients participating, you can only make a decent amount of money if the unit expenditure is high. So pyramid schemes in which you send, say, $1000 to each of five people, erase the top person and put yourself on the bottom--these are the schemes where distal recipients can still make some money. Of course, the risk is much higher.
The more accurate conclusion from your analysis than the foolishness of the participants is that you don't want all recipients to participate. And that's a realistic expectation, anyway.
I've seen analyses like yours--in fact, identical to yours--before. I think we need something a bit more extensive and rigorous. If you're a mathematician, you might want to consider such an effort.