Instead of looking at correlation between enneagram and socionics, why not use both systems to explain intratype differences. Differences within one enneagram type explained by socionics,
or if your prefer, differences within one socionics type explained by enneagram.
For other variables there could be correlations too, for example gender could be correlated to type: feeling types I guess more female and thinking types more male. Again with a distribution, no combination impossible. And each variable, like gender, is useful to explain intratype differences.