There are many implications of using data from other countries on LCA (in general, in any method of environmental accounting).
Examples. Using the energy efficiency of a very efficient country to represent environmental costs of energy production in a very inefficient country would result in understimation of environmental costs. Using production yields of crop production in a very industrial country to represent crop production in a rural country would probably overstimate productivity.
I think that this is not a very good idea when representing very different countries, regulation laws, technologies or production processes. Of course, there are countries in which the production processes are more or less the same (technology is the same, environmental conditions are very similar, etc.). Example. Probably, average production processes of milk in Holland or USA are more or less the same.
In the case of emissions, I think that technology or regulation could be important factors to use data from other countries.
Data availability is one of the biggest problems on LCA so very often practitioners have to use "proxy" data. The main problem with proxies is that your result will have a huge uncertainty and probably will be not representative. A approach would be to identify the contributions to impact and to variation. This will show you which are the most important processes. With this information you can try to characterize the data based on your local data. This is still a proxy but your veriablity and ucnertaitny will be known if not reduced. If you can, try to get ecoinvent 3, this database have the option of getting global values, which can be used with a know level of certainty.
We have developed the ACV Coclowen to avoid taking data from a indutrializado country. This method has two applications: a qualitative and a quantitative (the most expensive). Initially we have only mentioned as a Sima-pro software, and then apply the Coclowen compared, taking into account the different industrial contexts. Now we use the qualitative Coclowen, with experts from the areas involved and knowledgeable of the industrial state of Latin American country, especially Venezuela and Mexico, which is where we worked.