Rather than vaguely complaining that "My signal is poor in my house!!!" or "Why am I getting the slow speedtest.net result?", it is absolutely the better idea to pinpoint the trouble and deliver the message about it to VZW, Google, or Samsung. That's what the electronics user forum is supposed to do, IMO.
However, since the day one of GN sales, too many people in this and other forums have been mislead to focus on the wrong issues. First, number of bars. Second, signal strength display on the setting. Third, speedtest.net result. We now recognized that the first and the second doesn't really matter. How about the third? Do the speedtest.net results really matter? I would say yes and no.
Speedtest.net results, IMHO, have too many variables. You will see them when you repeatedly do the testings over time. On the same server, the same location, the same phone will show you highly variable results. On the same server, the same location, the same time, the same model phones also will show completely different numbers (my three GNs in my home did)!!! When I do the speedtest.net, I only trust the number when I do more than hundred tests over an hour period and average them. One, three, five, or even ten tests barely means nothing.
Many articles, posts, and thread which have shown up on the forum so far tell me that some (or all?) GNs are experiencing too early hand-offs as well as too delayed authentication when hand-off happens. This definitely could be a trouble maker for some GN users especially when they are at fringe areas. My current use is mainly focused on testing that part. Well, I may be right, or wrong. However, either right or wrong, I honestly believe that my focusing on pinpointing the trouble with my best knowledge will be the best choice to do, since it may help VZW (or Google or Samsung) make my GN a little better to utilize.