While the differences between memes and genes are interesting, I find that too much emphasis on the differences causes people to prematurely reject memetics - and then miss out on all its valuable aspects. So, I'm a little concerned that Razib may be inadvertently contributing to this problem.
Razib pointed out that gradual migration tends to result in the migrants adopting the memes of their new home - while their genes tended to survive the transition much better.
Someone who is sincerely looking for the corresponding area of genetics, should consider the genes in symbionts (parasites or mutualists) - not the genes in the humans. That is because cultural creatures (with memetic heredity) are much more like parasites and mutualists than they are like their human hosts.
So: do gradual human migrants pick up the colds, flus, tics, lice and bed bugs of their new home? Do they adopt the berries, grains and livestock of their new homeland? In some cases you'll find that the answer is "yes".
Also in the cultural case, sometimes the corresponding answer is "no". Consider the "surname" memes - they typically manage to survive the transition intact. Or imagine an inventor - some of their memes may survive and thrive in the new ecosystem - almost as if they had brought across a potent new viral strain.
From this broader perspective, memes and genes turn out not to look so very different. The dissimilarity arose from considering one type of meme and one type of gene. If you are considering memes that spread horizontally between hosts, the appropriate point of comparison is genes that spread horizontally between hosts - e.g. the genes in viruses.
Anyway, it is nice to see that Razib is at last using the "m"-word - after originally having some doubts on the topic back in 2004.