Agreed, which is why I don't think it is an accurate representation of the norm, and why you end up with all four couples in the 1st generation being MRCAs for the 4th.
No, I don't think that's the reason.
Let us define a "stem ancestor" as one having the following properties:
(a) S/he is an ancestor of everyone alive today.
(b) None of his/her children have property (a).
Now, intuitively, their ought to be quite a lot of these. It is possible to come up with family trees where there's only one, and even to come up with such family trees such that that there isn't a bottleneck: but it takes effort: such family trees are extremely factitious.
Now the question is: what are the odds that the most recent stem ancestors are contemporaries? (Of course, if they practice strict monogamy, the answer is 100%, and we should have to ask a slightly different question.)
Now the SkepticWiki diagram
is artificial in that it puts all the stem ancestors in the same generation, again, to make the diagram neater. However, I'm not sure that the small population size is significant.
I might do a computer simulation at some point and try to see how population size affects the odds, but the question has no real biological significance, so don't hold your breath.