For example, it makes the internet less safe - people remember just the company name foo, foo.x is the real site, somebody enters foo.y and lands on a scam site
and this is a problem why? i can see taking issue with allowing unicode tlds. where some russian characters are almost identical to a latin based char-set for instance, but if you can't tell the difference between bankofamerica.com and bankofamerica.cn, you probably also fall for a lot of other scams offline too.
Would you recommend a non-internet-savvy user from china to rather go to paypal.cn or to paypal.com/cn ?
What would a non-savvy user normally do?
In this case paypal.cn redirects to paypal.com/cn, but would you have known?
Another example: Python.com redirected to a hardcore porn site for years. Not that there is anything wrong with that, but I think both sites and their respective audiences would have been better off if there wasn't .com, .net and .org
I think there isn't going to be a system where the domain name points to the site that is more "legit" or "relevant", that is the job of search engines. Laypersons already use google instead of dns and enter "facebook" or "facebook.com" because they 1) don't have to remember the TLD and 2) get to facebook even when they made a typo.
So a domain name should probably still belong to whomever registered it first. I'm not a DNS wizard but using whois I figure python.com was registered one year before python.org, which would have become python-lang.com or something more descriptive if there weren't these other suffixes.
Except python.org is so much better than python-lang.com that isn't even funny.
Case in point ... I searched for "ruby language" dozens of times before learning "ruby-lang.org". And the ".org" suffix was the easy part because my brain associates those domains with non-profit stuff.
I also find TLDs to be useful for figuring out the target of a website. E.g. ".co.uk" / ".de" / ".eu" are usually companies with headquarters in the European Union.
