I've just been playing with Lost Cousins, another website/service dedicated to helping genealogical research. What is unique about Lost Cousins is that they match up names entered by subscribers based on more than the name, they use the detailed page reference from the 1881 census to disambiguate the surnames. This is really great, much increasing the chances of real matches, assuming that people correctly identify their ancestors in the 1881 census in the first place.
By comparison, I found that Genes Reunited had a lot of false returns, approximately 90% of enquiries about individuals in my tree were unrelated, and could be excluded very quickly on known information. In addition, manually searching for possible matches led to hundreds of potential matches, each one requiring page views to exclude. So cumbersome that I gave up - in all probability many other people give up too, undermining the whole purpose of the site. So I was interested to learn that Genes Reunited has just bought the 1901 web service set up by the PRO and Qinetiq, thinking perhaps they would do the same thing based on the 1901 census (arguably a better one to start with, in that family recollections are more likely to lead to finding individuals in the 1901).
However, it seems that Lost Cousins have entered a patent for their scheme, so it remains to be seen whether Genes Reunited will be able to improve their system on these lines. In any case, Lost Cousins have raised the bar in the "finding distant cousins" game.
