We already imagine that if Bitcoin gains a large pct of the world's txns, it may not scale, resulting in high txn fees, and therefore may end up being used mostly for large transfers. In your 1000 entities scenario, the only difference is that instead of large opaque transfers (we don't know anything about the BTC transfer just that it happened), we can see that X coins moved from SC A to SC B.
But running a SC does have overhead and risk. It seems obvious to me that companies will use the mainchain if possible (that is, it scales and can capture the transaction).
You can't stop innovation. If in the SC future there will be 1000 sidechains backed with Bitcoin, in the non-SC future there will be 1000 altcoins, and Bitcoin's market cap will be 1000 times lower then it could have been.
sure it can scale. Gavin has proposed at least one.
how can you see coins move from SC A to SC B? the Bitcoin network won't be monitoring all these SC's and there aren't enough mining resources to be able to MM all of them. on the contrary, they will be opaque and maintained by trusted owners of the SC's and as such are different ledgers than Bitcoin.
i don't want to stop innovation. i just want truly needed innovations to be implemented on MC. i disagree that it's so dangerous. we've done it before. maybe no one has introduced such a fundamental change b/c we haven't needed one? code has to be thoroughly vetted and tested on testnet or even on these federated server SC's. just don't introduce an offramp spvp to the source which breaks Bitcoin.