It is. SafeCoin is a type of StructuredData. Anyone can create their own types of StructuredData with the same capabilities as SafeCoin, except for the function of buying resources from the network itself.
It wouldn’t, you gain nothing by it, it only destroys the fungibility of SafeCoins. I’m not even sure the data field of the SafeCoin type will be mutable. Fungibility is a good argument to opt for immutability.
Limited to 100 KB, but you can point to other data on SAFE in that 100 KB, so it’s effectively unlimited.
This is incorrect, SafeCoin is not like Bitcoin and all. SafeCoin can divide far more than that in theory, but it has to be explicitly enabled in the code. This is because unlike a Bitcoin, a SafeCoin is an actual data entity. Bitcoins are just balances of addresses on a blockchain. A SafeCoin is a data entity (object) that has an owner field which contains one or more public key of the owner(s) of the data entity. It has to be recorded in the SafeCoin itself that it is divided into parts. This creates extra overhead, potentially a lot, so the current plan is to only enable divisibility when the SafeCoin market price warrants it. This is no obstacle to your concept though, because virtually unlimited asset tokens like SafeCoins can be created for whatever purposes we want.
Seems unnecessary to me, the SAFE network provides the same capabilities as magnet links and more.