Is Costa Rica more unequal than El Salvador?

I am currently in the midst of a research project evaluating the long term determinants of inequality in the eighteen Latin American countries.  Because I am interested in COMPARATIVE historical processes and there are only eighteen countries, econometrics analysis is not particularly useful.  Instead, I will probably use qualitative comparative analysis (CQA) as an organizing devise.  This requires creating sets of countries closer to the very unequal and not very unequal categories (since we are studying Latin America all countries are very unequal).

Here it is where the problems begin.  The data is not totally consistent with what I think we know about income distribution in Latin America.  Following Gabriel Palma's work, I organize countries in two criteria: the level of the Palma ratio (relation between the top 10% and the bottom 40%) and the middle groups (deciles five to nine).  I establish cut-off points that are related to global patterns of distribution.

There are two main problems with this data: one about cut-off points that we can ignore for the moment and a more relevant one (where any feedback is most welcomed) having to do with the various of specific countries.  The place of El Salvador and Peru (and to some extend the Dominican Republic) is particularly surprising. 

In particular, compare Costa Rica with El Salvador: despite recent changes in both countries, can we really assume that Costa Rica is more unequal than El Salvador?  Where are El Salvador's fourteen families?

Of course, this is not a problem of my research alone--which is the most worrying bit.  How much should we value quantitative research on inequality that is based on such questionable data?  Are we really measuring INEQUALITY when considering income distribution based on household surveys?