Were the raw data consistent? I can understand that the 3 companies were interpreting the raw data differently, but were the underlying snp frequencies consiste

Were the raw data consistent? I can understand that the 3 companies were interpreting the raw data differently, but were the underlying snp frequencies consistent? If they are not consistent, then the whole thing is pointless and they need to work on better kits. If they are consistent, it still doesn't say much. Genome wide association studies (GWAS) are very hard to control. And most papers tend to ignore external factors when they conduct the analysis. Do these group of people have this phenotype because of particular snps or do they have it because they have similar diets, similar geographic area, similar age range?