This is an incredibly complex question to answer. The interface between a transport and DAC vary from unit to unit. The combination of any number of cables, connectors and adaptors may move you closer or further from the "correct" termination impedance.
That being said, having the termination "wrong" may sound better, depending on your preference for tonal balance and perceived resolution. You answered your own question with your remarks about the Silver Link II's. Your experience is not uncommon when making this type of test.
In theory, the SUM of all the connections should be 75 ohms. The BNC was designed to be a true 75 Ohm connector, the RCA is not.
The errors of the cable and connector together may balance each other out, making two improper impedance, correct.
A few years ago Audio Research manufactured a RCA to BNC adaptor. My experience with that adaptor compared to a direct connection (removal of adaptor) was about as exciting as your Silver Link test.
In the end, you are doing the right thing in experimenting.
That being said, having the termination "wrong" may sound better, depending on your preference for tonal balance and perceived resolution. You answered your own question with your remarks about the Silver Link II's. Your experience is not uncommon when making this type of test.
In theory, the SUM of all the connections should be 75 ohms. The BNC was designed to be a true 75 Ohm connector, the RCA is not.
The errors of the cable and connector together may balance each other out, making two improper impedance, correct.
A few years ago Audio Research manufactured a RCA to BNC adaptor. My experience with that adaptor compared to a direct connection (removal of adaptor) was about as exciting as your Silver Link test.
In the end, you are doing the right thing in experimenting.