I incorporate several verification points in a single test method and the reasons are as below:
Before performing any action on an element I would first like to verify if the element is present on the page in the first place. There are some possibilities that the UI control is renamed or removed. The same logic applies for pages as well. I verify at every step if I am on the right page before performing any testing by verifying the title of the page. I thus have several verification points in a single test method. I first verify if I am on the right page, then whether the required UI control is present and after that I perform the intended verification point of the test case.
Certain verification points can be logically grouped in a single test method. For example, when I test "post comment feature" on facebook I have got two verification points. One is to verify if the comment is posted successfully and another is to verify if my friend can read the comment that I just posted. It makes sense to group these two verification points in a single test method since they are logically related. It is incorrect to claim that post comment works if others can't read my comment.