Why did Gallup even bother?

On this page we find the results of a post-debate poll conducted by Gallup. I really have to question the reporting of these numbers. At the bottom of the report we find the following:

All results are based on telephone interviews with 615 registered voters, aged 18 and older, who watched the presidential debate Sept. 30, 2004. Respondents were first interviewed Sept. 28-29, 2004, when they indicated there was some chance they would watch Thursday's debate and were willing to be called back. For results based on the total sample of debate watchers, one can say with 95% confidence that the margin of sampling error is ±4 percentage points.

In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of public opinion polls.

Polls conducted entirely in one day, such as this one, are subject to additional error or bias not found in polls conducted over several days.

So not only are these numbers affected by the typical sampling error (which is typically between 3 and 5 percent), but they're affected by "additional error or bias" because they were done in such a short amount of time. Apparently Gallup doesn't know what that error percentage would be since they didn't report it. It could be anything! So why bother? Is Gallup jumping on the "report it no matter how suspect" bandwagon with C^BS?