It's good for cryptocurrencies in general that Bitcoin can innovate and remain relevant. Many people have viewed Bitcoin as THE currency of the future, many people believe that Bitcoin's first mover advantage matters and that it will be capable to evolve and adopt cool new features. That is the reason Bitcoin's market cap is that high compared to all newer coins. If Bitcoin gets outcompeted in the long term, that should be because of inherent economic flaws in its philosophy, not because it lacks a killer feature X or because it has slow confirmations, etc. If Bitcoin fails due to lack of innovation, it can seriously undermine public image of cryptocurrencies and give impression that every cryptocurrency becomes obsolete in a few years. It's important to insist that there is place for economically sound alternatives to Bitcoin, it's important to insist that Bitcoin's monopoly is problematic, but explicitly having the goal to become bigger than BTC and make BTC obsolete isn't the best approach.

Unfortunately, .name is already in use since 2001:
http://en.wikipedia.org/wiki/.name

There are ICANN applications for .free:
https://gtldresult.icann.org/application-result/applicationstatus/viewstatus

.key looks pretty catchy and makes sense.

Using two-letter name is problematic since those have always been reserved for ccTLDs.

Storing names in UTF-8 can cause compatibility problems, it's safer to require that all names including non-ASCII characters be converted to punycode first. Many applications and standards assume that Unicode domain name is equivalent to its punycode representation, and BTS DNS resolvers should always treat them as equivalent. Spoofing names using homoglyphs can be a problem; traditional TLDs normaly solve that by restricting allowed characters to one or several writing systems, BitShares DNS is for global use and can't easily solve that (I think Firefox shows domain names in punycode when they are not considered safe).