I'm not sure you fully understand how polls work... If the poll was taken over the internet, promoted through, say, something like Google AdSense which would make it turn up with higher prevalence on sites such as macrumors.co.jp, then the results can not be generalized in the way you describe.
So, I was in full agreement with you when you were saying you can't draw any conclusions from this poll. Where I started to disagree is where you started to draw conclusions from this poll...
I'm not sure you fully understand how polls work... If the poll was taken over the internet, promoted through, say, something like Google AdSense which would make it turn up with higher prevalence on sites such as macrumors.co.jp, then the results can not be generalized in the way you describe.
The fact that this was an internet poll makes it suspect to begin with. The fact that it was an internet poll that only pulled in 400 responses makes it pretty much useless.
To the point that this posting should probably be called "Useless Survey"...
So, I was in full agreement with you when you were saying you can't draw any conclusions from this poll. Where I started to disagree is where you started to draw conclusions from this poll...
Personally, my guess is that the iPhone will be somewhat popular among foreigners (especially Westerners) living in Japan, but not so popular among Japanese people. Even die-hardMac & iPod fans are mostly giving the iPhone a pass here.
If you read some of the early reports from the Japanese press, one of the things they are all angry about is the fact that the iPhone has handwriting recognition for Chinese characters when in Chinese language mode, but won't do it in Japanese mode (even though they use mostly the same characters). Japan and China are like fierce "sibling rivals," and giving a very useful feature to the Chinese but omitting it for the Japanese without any explanation is like a slap in the face to a lot of Japanese people.
The lack of MMS, which is heavily used by all the other Japanese celphone providers, as well as the lack of video calling (not super popular here but a fairly important checkbox feature to have), Felica/"Saifu Keitai" (aka Wallet Phone, RFID money card/train pass), and 1seg TV (mobile HD broadcast offered by all major TV networks), results in the iPhone comparing poorly to even the free throwaway phones you can get from any of the Japanese providers.
If only Americans would do this more often, the additional competitive pressure might lead to better products. When we base our purchase decisions only on ads, without looking at what we are getting, we often aren't getting the best product for the job.
A poll's error margin is related to the population being studied. In this case, the population being studied is not "Japan", it's some subset of Japan. We don't know what subset. If there is any correlation between the selecting criteria for the subset and the response (people who clicked on "win a free iPhone" for example, or people who visited the Au forums) then the results are almost impossible to generalize. You'd need to know how to convert the likelihoods of a response from the sub-population to the larger population which might be achievable through good historical data if you can show the relationship has been consistent enough over time-- but no such historical data can be available for a product this young.
It's not what you'd expect. It's not competitive pressure, it's a checklist. When you sit down to design a new product, you pull out the top three last generation products and make a list of all their features. You take the union of those sets, add a feature or two for uniqueness (which will then become standard across the board next generation) and make generational improvements on everything else.
