The way this kind of mechanism, the entire concept of recommendations, is largely by volume, which is what lends itself to the network's capabilities. As opposed to asking an individual, who has a low chance of necessarily knowing the one piece of software that does what you need, you ask the group of individuals in the hopes that the one who does know is amongst the group. Extending from that, the people who excel within this group are the ones who can provide the most help, by a matter of knowing the most.
Whereas in programming you become an expert by knowing the most about the field, in here you become an expert by having the largest breadth of knowledge with regards to solving problems. Much like programming has languages that people work in, recommendations have domains of problems that are solved. How one hones that experience is also largely the same - some people will specialize in particular kinds of tasks (say, a particular user might be an expert on all things used for the many different tasks of home management), others don't necessarily specialize and instead work off of the things they run into on their daily movements. Some will find usable solutions and work with them, others will constantly revisit what they know and see if better exists out there. We all build experience with each thing we try that solves our own problems. So to a point, the greatest experts tend to be those who have needed to solve the most problems on their own, as that would have lent them the necessity to investigate as many problems and solutions to be able to share.
This isn't, ultimately, something people necessary work towards in an active sense. This is largely a passive expertise that is built up as one goes through life (though, again, some people do strive for more). That doesn't make it any less respectable than other kinds of expertise. It's simply a different case.
That, I feel, is what it would mean to be an expert in this site. There are the people who may not know things off-hand, but are experts at research and testing and so they can provide extremely quality answers. I'd though say that that person isn't an expert at recommendations as much as they are an expert at research and testing. Which does indeed help, though.