I'm also a fan of LT, but most importantly I'm a fan of spreading the wonderful feeling you get from reading. If LT's recommendation algorithm is only good because of its pool size, then it's bound to fail at some point in the future. Rather than delaying this inevitability -- which effectively sweeps the issue under the rug -- the better solution would be to ensure that their recommendations stay high-caliber regardless of scale. I'd imagine the only way to perform this test would to be for them to actually grow in scale, so they should be advertised as much as possible.