This is basics for sure but might make for a good cross check for versionitis.
The server the services is running off of is a 10.4.1 server. "Enterprise" was the big change at 10.5 that didn't really have a lot of its kinks (portal/server) refined until 10.6.1. Obviously components of R&H have had various changes along the way.
Your working scenario falls in that early "Enterprise" changeover and isn't far removed from the 10.4.1 service. I'd be curious to know if you pushed the same service to a 10.8.1 or at least 10.7+ environment and pulled it into the 10.8.1 Event Editor if you saw the same results (15 min) or normal (1 min)?
Do you have other services coming from this same 10.4.1 server into the 10.8.1 Event Editor that load with normal expected load times?
Only other item I'd question just from a cursory consideration of timelines, where does this service live (on-prem, cloud, etc.) in comparison to the server hosting 10.5.1 Event Editor vs. 10.8.1 Event Editor?