Unable to Join Known In-Progress Duty

[Jekyll]

Time: 11:34 a.m. EDT
Frequency: Multiple Times
World name: Sargatanas
Character name: Henry Jekyll
Class/Level: DRK/60
Party or solo: Solo
In-game time: ~3:00 a.m.
Steps:
1. LS member requests assistance with in-progress duty (Library HM) due to tank dropping out
2. Select in-progress and queue for Library HM
3. DF pops up with new instance
4. Repeat 1-3 3x's
5. Get locked out of DF due to penalty
6. 20 minutes later, LS member still reports no tank in the instance

Connection Specs
- Type of internet connection/provider: Fiber Optics/Fios

This has occurred to me at least 3 times since 3.4. While it may seem like anecdotal evidence, I've yet to be able to join a "known" (using the word known because I know it is out there due to LS or FC member being there) in-progess duty as the needed role. It ends up not getting filled for quite some time (20 minutes or more in the latest instance).

[Mholito]

Are you sure your language selection matches the language settings of the known instance?

[Jekyll]

Are you sure your language selection matches the language settings of the known instance?

This could be the case. I only have E selected as my language as I've found it an improvement for my queue times (NA data center and all). I know at least one of the 3 people I've been unable able to help does have all languages selected. So assuming one person in the group had only one of those, they could've ended up in German, French, or Japanese instance.

I believe that they changed this some time ago so that a new instance is ALWAYS prioritised over an ongoing instance. That means this is working as intended, unfortunately.

I'm pretty sure that is not the case... I've dug around through patch notes. This wouldn't make sense.

[Culfinrandir]

I believe that they changed this some time ago so that a new instance is ALWAYS prioritised over an ongoing instance. That means this is working as intended, unfortunately.

[Alistaire]

I believe that they changed this some time ago so that a new instance is ALWAYS prioritised over an ongoing instance. That means this is working as intended, unfortunately.

If this is true then the bug is with whoever would design such a backwards thing.