It's obvious that the idea has alot of potential (as well as technical hurdles, hence the distinct lack of Bitcoin sidechains in operation since the idea was first written up)
I like the experimental possibilities; Ethereum's EC20 protocol (?) could be viewed like a sidechain system, where many different tokens (that do not necessarily have to be currencies
) are run in parallel with the underlying Ethereum blockchain (Buterin always said "Ethereum is not designed to function as a currency", and to a certain extent that's true, you're buying shares in the overall platform when you purchase ETH). This direction was originally fulfilled with the "colored coins" approach in Bitcoin, and at least one noteworthy success (Omnilayer's hosting of the Tether stablecoins) has been using Bitcoin's main blockchain for this secondary purpose.
Having a sidechain for an internet domain name system, cryptographic shares in companies (even other financial contracts), or just for serving and querying keys (PGP system could use such a sidechain in light of this summer's keyserver flaw) could all be very aptly implemented as Bitcoin sidechains.
There's a problem with all the possibility though; freedom to experiment includes freedom to design deceptive or disastrous sidechains, and the corresponding freedom to make unwise decisions about using such chains. I expect that a breakthrough in Bitcoin sidechain tech and a subsequent explosion in marketing of new sidechains is going to lead to alot of fraud and losses of over-excited investors.
No-one's holding your hand to bail you out if you screw up, permissionless capitalism means
all trades are final, even if
"it isn't fair". Buyer beware.