"I'm wondering if anyone has any idea of the best way to load this and extract the genotypes from the region I need"

See tabix

Tabix indexes a TAB-delimited genome position file in.tab.bgz and creates an index file in.tab.bgz.tbi when region is absent from the command-line. The input data file must be position sorted and compressed by bgzip which has a gzip(1) like interface. After indexing, tabix is able to quickly retrieve data lines overlapping regions specified in the format "chr:beginPos-endPos". Fast data retrieval also works over network if URI is given as a file name and in this case the index file will be downloaded if it is not present locally.

I am wondering what is the difference between this file (ALL.wgs.phase1.projectConsensus.snps.sites.vcf) and those name by chromsome?

"sites.vcf" doesn't contain any data about the genotypes ;)

by the way, are you experiencing too right now some .tbi files "incorrect" data? (90 bytes as size is way too small...)

You can try use vcf2geno http://cran.r-project.org/web/packages/vcf2geno/index.html It takes tabix-indexed vcf file and extract genotypes for you.