use the most selective index

[conker84]

After the #99 we need to manage the following use-case:

if we have two Neo4j instances that are configured as Source and Sink with the CDC schema implementation we need to select the most selective index in order to speed up the import process.