There's a subtle difference between SID and SERVICE_NAME and I think the code should somehow account for the difference. Here's why... An SID points to a physical instance at the database level (Service Identifier). If you ask for a SID you only get one shot at the connection (and hopefully the database is up). That's the legacy method of connecting to Oracle.
However, for environments that have HA installations of Oracle using RAC, the SERVICE_NAME is what is used to point to a database instance (instead of a physical Service ID). If a given database instance is down and you're using SID - meaning you're looking to connect to a specific database, the connection fails. If you're in an environment with multiple instances of the target database, then RAC figures out where to send you if you're using SERVICE_NAME and you will land on an useable instance. This is the typical way to connect when Oracle is load-balanced and in a high-availability environment.