Unify the way how we read configuration in Enceladus

## Background
The way how application configuration is read in `spark-job.jar` is a mess. Sometimes it's our own `ConfigReader` sometimes the typesafe config. There are same functions defined for it, sometimes as implicits sometimes as locla ones. Needs some unification

## Feature
Refactor the code to use only `ConfigReader` to access application configuration.
The only exception being Hypedrive plugin, where it's to be considered.