GeoMLA vs cokriging

[influence-of-canopy-on-temperature]

Hi,
I was wondering whether the RFsp could also be extended to cover coKriging problems, such as if there is another point database (points in different locations) with potential covariates (http://www.css.cornell.edu/faculty/dgr2/_static/files/R_PDF/CoKrigeR.pdf). In this case, I guess we would need to make buffer distance layers both for the predicted variable and for the covariate(s), don't you think?

Best regards,
Vitor Vieira Vasconcelos