Some trial and error showed me the way to remove the gap, is to set the same altitude restriction on both waypoints. As soon as there is any difference, this gap is introduced.

In Europe, there are quite a lot of examples where the published altitude restriction on the same waypoint, when in the START and when in the approach, is different. I know that doesn't make a lot of sense, but there you go. As a result of that, a gap is always introduced and can only be removed by tweaking the restrictions manually.

Although I understand the logic behind this heuristic, I believe this is non optimal behaviour, as an unconnectable gap means you need to remember to manually sequence to the second instance of that waypoint when on the leg to the first instance, without any warning or prompt to say this is required.

A better choice would be not to introduce this gap in the first place, or at least have some logic that forced the pilot to acknowledge or act on the gap.

Can this be considered for an upcoming release?

I'm wondering if this altitude data conflict is the reason... must the gaps be eliminated before an approach can be activated even if you've flown beyond the gap?

Check  kmia-- TTYLR4.TTYLR to 8L

It's not clear from your post whether you are only concerned about the IFD auto-activating the approach, or if you are also unable to manually activate the approach as well.  

Interestingly, this only happens when an arrival is loaded.

Am I missing something? or is the box faulty?

In both cases, with the arrival loaded, sequence to approach does not seem to work.

This is normal gap behavior for the IFD. The “Activate Approach” LSK 4 hook will show up on the FMS FPL tab when the IFD active waypoint is followed by a gap before the approach. On the PALMZ2.PALMZ for 8L arrival, the Activate Approach LSK 4 hook shows up upon crossing COOFS when DHP becomes the IFD active waypoint.

With all FMS’s I have used, some pilot action needs to be taken to get the FMS to navigate to an active waypoint on the other side of a route discontinuity (gap). In the IFD’s case, all “Activate Approach” is gets the IFD active waypoint to cross the gap between the arrival and approach by performing a direct to the first point of the approach (PABOY in the case of KMIA LOC 8L). There are other ways to get the IFD active waypoint to cross the gap without waiting for or using the Activate Approach LSK. For instance, if at any point on the PALMZ2 arrival, ATC instructions are to proceed direct to PABOY, just perform a direct to PABOY on the IFD using the direct to button. The IFD active waypoint is now on the approach side of the gap and the IFD will give proper guidance for the approach. Alternatively, if ATC instructions are to fly an assigned heading to intercept the LOC 8L final approach course while on the arrival, either an Activate Leg to LICEY or a direct to PABOY with a subsequent setting of an OBS course of 092 to PABOY would get the IFD active waypoint on the approach side of the gap with subsequent proper approach guidance. Essentially, the “Activate Approach” LSK is not at all necessary on the IFD like it was on the GNS system.

One last thought on this scenario, if an Activate Leg to LICEY were performed, the IFD will only display the magenta active course line as far back as PABOY; however, the IFD will generate proper CDI guidance for that course if the intercept actually occurs prior to PABOY even though no magenta course is displayed for that area on the IFD map. If after final approach course intercept in this scenario, the aircraft was still outside of PABOY and either a magenta course line was desired outside of PABOY, or the pilot wanted PABOY as the active waypoint; a direct to PABOY could then be performed.