I went so far as to google the quoted question. It's nowhere, not at Harris, not in the MSM, though the phrasing was quite similar to that used by a writer in the Conservative Review. So, yes, it could be an interesting way to do a psychological study on messaging, but if so, I can't find any evidence such a poll was conducted.
Actual polls, for one thing, would separate out those long sentences into several, each asking a distinct question. For example, one could ask in a poll if one believed in standing up to China and Iran. I would estimate at least 80% of any polling sample would answer "yes" to that one. As would pretty much any candidate for national office in the United States.
I think Needs covered this pretty well, so I won't go on.