One timezone covering enough longitudinal variation to want five timezones is a dumb idea. Doing the same thing in a country with enough east-west span to cover ten timezones would be an incredibly dumb idea.
In your system, standard store hours have become a random regional fluctuation instead of a strictly maintained standard. I still need to maintain timezone information, but I ALSO need to do math in my head to figure out if it is an appropriate time to go to a certain kind of store.
There is huge variation in these sorts of things worldwide. See, for example, http://www.spanish-town-guides.com/Opening_Hours.htm
(PS Spain has multiple timezones: http://www.timeanddate.com/worldclock/city.html?n=141 vs http://www.timeanddate.com/worldclock/city.html?n=683 for example)
o Restaurants open at 11:00 AM.
o You check out of your hotel no earlier than 11:00 AM
o You can check in at 3:00 PM.
But, I don't think I've ever been to a country where the checkin/checkout/restaurant opening wasn't generally true.
Admittedly, I haven't been to the "One Time Zone China" - that might throw my theory out the door...
Restaurant opening hours vary dramatically by country, largely due people tending to eat at very different times in different countries. To pick a restaurant I walked past yesterday: http://www.restaurant-thierry.fr/index.php?p=contact — open 12pm–3pm and 7pm–11pm. From my experience here those lunchtime hours are fairly common, although more restaurants would open earlier than than in the evening (unlike Spain, where lots of restaurants don't open until at least 9pm or 10pm at night).
On the assumption that #2 is "no later than", then where there's consistency at all it would be noon, rather than 11am.
Check-in has almost no consistency at all. I'm currently in France, and lots of the hotels I've been looking at have a 6pm earliest check-in. Many others are noon or 1pm. I haven't seen one yet that's 3pm.
You're making an assumption of a standard time of day when stuff's open (I agree, 11:00 is a reasonable assumption). I can, however, make that same assumption. In your travel model, the traveler must know the timezone (and work with assumptions from there). In my mode, the traveler would need to know the “standard daytime hour” and work with assumptions from there.
So, both require one piece of info when traveling, but a single timezone wins out when synchronising.
Either way, you have to adjust something. Right now, you adjust your known time difference (i.e., change the clock, either by knowing the difference or by asking someone local). With that proposal, you would adjust the known opening times.
Why is adjusting the clock numbers better than adjusting local practices?
Why does a single time work well for China... and the US armed forces globally?
Shouldn't we be optimizing for long-distance collaboration, now that we have the telecom technology to do so?