Why not 100M ? It would be better comparable with BTC and satosis.
I didn't cover my thoughts on 10M in the post but here they are:
This is really about 2 things, making a liquid BTC-paired market possible and trader psychology. Firstly, I think using anything but a power of 10 would be a mistake, so 1, 10, or 100 million atomic iota to IOTA units should be considered.
1M would be a lower bound possibility in my mind, giving a supply of 2,779,530,283 decimalized units. But there are issues with coins trading in the hundreds of satoshi range. This is my own opinion but maybe it's shared. For one the price doesn't seem substantial enough and feel cheap. For USD/GBP/Euro-imprinted traders it can be hard to translate a price of 1000 satoshis into what they're used to using in day to day life. 10M on the other hand would be priced near USD cents.
100M would be an upper bound, giving a total supply of 27,795,302 decimalized units. I don't believe this supply is enough and would lead to the tokens looking too expensive. I could be wrong about this, and perhaps perceived scarcity could have a positive effect on price. But I don't think it would help liquidity.
10M carves a good balance between the extremes. The total decimalized supply of IOTA ends up being 277,953,028.3277761. The price (at current OTC IOU rates at least) would fall at about 0.0001 BTC/IOTA. This leaves plenty of room for the token's price to go up or down and still have a functional liquid market. It has enough value to not seem borderline worthless for 1 like NXT or XRP does. Its value could be more easily estimated in cents or pence for the bulk of us who are accustomed to thinking of monetary value in those terms.
