I use a C5050 with PT-015. I bought a DS50 a few weeks ago with the TTL sensor and tried it out on a dive trip to Lembeh Strait last week. After a couple of dives I gave up on the strobe as the photos were always way underexposed. 
After I got back I did a few test shots on the table at home, and a few results are at
uk.photos.yahoo.com/bc/d3m2m
(look at the "DS50 tests" folder; and BTW they are not in numerical order on the page so check the number).
The photos are all identical setup except with differences as follows:
1. Internal flash set to compensation -0.3 (no particular reason); DS50@off
2. Internal @-0.3; DS50@TTL
3. Internal @-0.3 but blocked; DS50@TTL
4. Flash off
5. Internal @Slave 1; DS50@off
6. Internal @Slave 5; DS50@off
7. Internal @Slave 1 but blocked; DS50@TTL
8. Internal @Slave 5 but blocked; DS50@TTL
9. Internal @Slave 5 but blocked; DS50@Manual
I expected the results of 1,2,3 to be substantially the same (ie TTL should produce the same exposure with any combination of internal strobe and DS50 on or off) but obviously that's not what happened.
I suspect what is happening is that the C5050 internal flash is varying intensity rather than duration of the flash to achieve the desired exposure, and the amount of light dumped by the DS50 in TTL mode depends only on the duration. I draw this conclusion partly by comparing photos 7&8 which appear to be identical; however it's possible that the operation in slave mode is based on intensity whereas the operation in slow sync mode is based on duration, in which case there must be another problem.
On the other hand maybe there's a problem with the particular TTL sensor unit or the strobe. I have sent an eMail to Ikelite asking about this, but since that was only a few minutes ago I'm not expecting a response yet ...
-David