Hello. There are a couple issues with the FMS that make transitioning from an arrival to an approach more difficult than it should be, coming from the perspective of a type-rated CRJ airline pilot.
The first problem arises when programming a STAR and approach procedure into the FMS at the beginning of your flight. When you're selecting the transition you want for the approach, the FMS gives you the option of vectors or one of the initial approach fixes for the procedure. If you select vectors and look at the legs page, you'll see the transition sequence as: [final fix on the arrival], DISCONTINUITY, [intermediate fix on the approach]. This is correct and how it should be. If you select one of the IAFs, the sequence becomes: [final fix on the arrival], [initial approach fix]. When the autopilot follows this sequence during the flight, it will fly direct from the end of the STAR to the IAF for the approach. This is not correct. There should be a discontinuity between the end of the arrival and beginning of the approach. This discontinuity is very important, as you don't want the plane to fly onto the approach before ATC has given instructions or started vectoring you. You might ask, "why not just use the 'vectors' transition option then?" It's because you might want to intercept the final approach course from further out than the IF. For example, look at the chart for the ILS 12L into MSP. If I select vectors as my transition, the legs page only shows approach fixes starting at HAMML. If ATC cleared me direct to ALGIN, or wanted me to intercept an extended final before HAMML, I wouldn't be able to do it.
For the second problem, I'll refer you to this thread for the FSX/P3D version of the CRJ, as it's basically describing a similar issue. We use the INTC CRS function on the legs page almost every flight when setting up for an instrument approach. Most commonly to create an extended centerline for the autopilot to intercept. Using the above example of the ILS 12L approach, I would move CCJAY to LSK2, enter 121 into the INTC CRS, and execute, creating an intercept course that extends before CCJAY. The current FMS doesn't seem to be able to do this. If you try to modify the INTC CRS when sequencing a fix, it gives a "NO INTERCEPTION" error. This makes vectoring onto an approach much more difficult.
Hopefully my description of these issues was clear. Let me know if you need more examples or pictures to explain it.