"Google actually tested this, you didn’t. So while the reasons you listed might be valid and a good reason to scrutinize the redesign with the literature in min

"Google actually tested this, you didn’t. So while the reasons you listed might be valid and a good reason to scrutinize the redesign with the literature in mind, I have a very hard time believing that Google didn’t do just that." And yet it moves! (http://en.wikipedia.org/wiki/E_pur_si_muove! http://en.wikipedia.org/wiki/E_pur_si_muove!). Google "testing it" isn't an iron clad guarantee that the feature doesn't suck. Anyone who has worked in a large company can well imagine how "tests" can support predetermined positions of higher management (say). I suspect the reason is simpler - these days Google is fat and slow and infested with mediocre MBA types who love nothing better than to "launch an initiative" like this. Never attribute to malice etc. Irrespective of how well Google tested this, if enough users say "And yet it sucks", then that's that. You could argue about how many users are "enough", but "Google tested it so they have to be right and everyone else is wrong" is a very fragile argument.