The risk of not holding monero is infinitely bigger than that of holding it.
With not holding, you can lose a sum equal to 1000x your net worth.
With holding, you can maximum lose the amount invested.
Sounds like Pascal's wager. That can never be the reason to do it.
I may be strange in that it seems perfectly valid to me. I pay taxes, because by paying, I lose only the sum invested, but by not paying, I risk unforeseen consequences. With Monero there is at least some upside, unlike in paying taxes which is a genuine lose-lose...
Yes, but the same is true for any other altcoin (even the most junky ones),
take the following assumption:
1.) you cannot invest
2.) you can buy a basket which is filled in an equal distribution with the top200 decentralized currencies
what would you do?
the same is true for the existence of any god. This is just not how logic works though.
even following the wikipedia article - pascals wager is not
a proof of god - I always understood it that way: even if there is a tiny possibility for an infinite outcome, it is rational to take the bet . I see nothing wrong with that.
If I follow your idea correctly you say that if you observe 1000 ravens, and they are all black you can neither conclude that all ravens are black nor can you conclude that the next observed raven is black. This is obviously true; but don't you think you can at least partially conclude that the next raven is black?
To come back to xmr and I usually do not take much out of ristos predictions: but what he is essentially saying is that he has observed 1000 ravens which were all black, and xmr is the 1001 he is seeing; his estimation of the likelyhood that this raven is being black is quite high.
Let's say there were originally 10 black ravens, and you concluded that every new raven would
Not be black, you would be wrong upto the point of there being 1,000 ravens(990 after the original 10 or 990 times you were wrong). To continue concluding that the next raven would
Not be black would mean that you're insane in Einstein's definition of the word.
It's overwhelmingly logical and rational to estimate that the 1001th raven would be black, given that the previous 1000 were.
However, there is still a chance that the 1001 raven would
Not be black, though the odds are 1000 to 1.
(given that you abstracted away the actual reasons why XMR is not "black" here but alas) 
