Ah, thanks. You point out two things which are probably overseen and forgotten by most, me included. If the cavity behind is sealed, the gaps around the speaker should result in a Helmholtz resonator at some frequency. (I believe it is good then when I have 10-30 cm of fiberglass insulation in the cavity between cabinet and the outermost concrete wall. Should provide some damping anyway.)

The second one; cabinet vibrations versus mid / tweeter driver, was given some thought in the construction of the DIY cabinets, though not as part of a flush mount detail with immobile cabinets. Thanks for pointing it out. I have no means of really measuring it but made the same crude “experiment” again and played at 95 dB / 1 m. With fibertips slighltly touching the aluminum waveguide for the mid / high, vibrations are hardly noticeable. I can only say they are lower than the cabinets. (A reason may be the use of viton o-ring seal between aluminum frame and the cabinet. I went through some headaches when the DIY cabinets were made and when I came across a NASA report saying viton tops neoprene for damping vibrations of audible frequencies at room temperature, I decided to go for viton as sealing element between driver frames <=> cabinet and machine screws through slightly oversized holes. The cabinet baffle is a 2 sheet constrained layer damping construction. On the aluminum wave guide, inside the cabinet, 4 mm adhesive bitumen sheeting was applied as vibration damping.)