Depends on the population you're interested in, but simply said, yes there is overlap to some extend between the HapMap samples and the 1000 Genomes samples.

I know that for the CEU samples there is a mis-match between the HapMap genotypes and the latest 1000 Genomes release. Though, for the CEU samples the idea is to get the missing ones sequenced within the 1000 genomes on a later stage as well (depending on sample quality etc).

As for the merging I don't really understand what you mean. The 1000 genomes genotypes is actually a superset of the hapmap genotypes. So genotypes called in hapmap are also called in the 1000 genomes so there is not much to merge (or do I mis interpret your "merge" question?).