I only posted in that thread because I was linked to it because it's goofy - I don't read this forum usually because I have other things to do in life. So the sample is self-selecting and skewed towards people who are interested in this particular issue. If we assume that people who don't post on the forums are generally happy with the direction the game is going, we can assume that the people who have posted in that thread are going to be the ones who care about the issue, and the people who didn't post, the vast majority, would have been fine with instant ship transfer.
Statistically speaking, none of this is true. For a population of 2 million a random sample of 2600 would give us a pretty good confidence level, but this wasn't a random sampling, it was self-selected. So no, you cannot have the confidence level you are expecting. Test design is the critical difference between statistical masturbation and actual research.
What Frontier has done is made an "official" poll using two extreme options because they couldn't use the self-selecting poll made on the forums or gather useful information from it. What they have failed to do is make the poll they have created here actually useful because most people do not follow through with polls that are emailed to them or bother to vote about issues that they do not care strongly about. So most people would likely be very happy with instant transfer, but will not vote because they trust Frontier to make the best gameplay decisions possible when designing the game (though I don't know why they have that faith at this point).
Because most people trust the game devs to design a game that is fun and cool and not lame and boring, they don't bother voting in polls about gameplay unless they are of the sort that is afraid of fun, cool gameplay because they use Elite as a reality-substitute rather than an entertaining video game. To the latter, it is very important that Simguru Braben not make a game too fun because if they are having fun, they will realize that it is not actually their real life, which is joyless and tedious as all real lives are.
Basically, the statistical methods being employed here aren't actually meaningful and shouldn't ever be used to influence actual game design, instead, simple tests should be employed, like "is it fun when we actually play it."
Sandro straight up said that they have tested it both with and without a delay, and a delay added nothing, when they actually played it. People who have not played it both ways are saying they would rather play it the not-fun way, but they cannot know that because they have not played it either way. Instead, they are worried about their immersion, which is a silly thing. I would rather gameplay decisions be made by people who have actually playtested it and not by biased polls.