The value of changing the FMA depends on the viewpoint.
A conventional view of automatic flight is that the aircraft's state is determined by using the basic instruments; if the flight path is not as required then the auto flight system should be adjusted (or disengaged) and the modes and selections cross checked with the crew's intention.
An alternative view reverses this, where the FMA is used to deduce what the automation is doing, or should be doing, with little reference to the basic instruments. Then the aircraft flight path is inferred from the annunciation. This often results in 'what is it doing now' and contributes to reduced awareness of attitude or speed.
Locating mode annunciation closer to the relevant instrument parameter is an interesting idea, but it might not direct attention to the associated instrument parameter as assumed.
If a pilot looks for a mode annunciation, s/he will 'see' that mode without necessarily being aware of the adjacent instrument display; vice versa, look for an instrument reading and not 'see' the annunciation.
Whilst many accident and incident reports cite 'failure to look at the FMA', such conclusions more often hide problems associated with lack of awareness of the aircraft state, poor instrument scan, the 'mantra' need follow SOPs and monitor everything, but with little thought of workload and the importance of understanding.
The proposal is interesting, but is it a solution seeking a problem?
What specific issue is it designed to address?