Karen, I am in total agreement with you and seem to be using some of the same methods that you mentioned, i.e., switching to the pedigree view, largely ignoring the shared surnames list, etc.
I wonder how many people simply leave the page of their match without ever bothering to take a look at their actual tree or how many never bother to look at a distant or low confidence match?
After doing this for a few months I have learned that about a half dozen or so of my ancestral surnames are shared with those with whom I ultimately discover to share a common ancestor with. Now in order to save time, once I open the pedigree view of a match, I will simply insert those most likely surnames into the "find a person in this tree" box and many times I will locate our common ancestor quickly.
I also would never hesitate to view someone's tree simply because it has a large number of individuals in it. Sure there may be a high degree of cutting and pasting at work and much of their earlier generations may be lacking documentation but in my own case since I have very few persons listed from the pre-immigration days, I'm not expecting to find any common ancestors from any earlier times anyway and just because someone's earlier generations may be inaccurate or unprovable doesn't mean that what they show for the more recent ancestors isn't accurate.
I have also been tested at FTDNA, as have both of my parents. While the Ancestry.com interface could certainly benefit from having some of the filters available on FTDNA, I was still able to confirm more common ancestors on Ancestry.com in one week than I did in over a year's time on the other site.