A dyvbuq has taken possession of the bride: on dagesh hazaq in bg"d kf"t letters

This one is irremediably geeqy. Apologies to the nondiqduqgeeqs who dropped in.

When there's a dagesh hazaq, the degushah letter is geminated--it is repeated, as though there's a sheva nah between the two occurrences of the letter. In Professor Moshe Goshen-Gottstein's Diqduq ha'ivri hashimmushi (Tel Aviv: Schocken, 1996), two examples are given (p. 28):

עַמִּים = עַמְמִים and שִׁבֵּר = שִׁבְ-בֵּר

The first of these is a little problematic, since Professor Goshen-Gottstein just told us (p. 26) that a sheva between two identical letters is a sheva na, not a nah. Maybe a hyphen, as in the second example, would have helped. But we know what he means, and the first example isn't the one I'm really interested in.

I see the logic in the second example, and it may well be theoretically correct. But assuming you pronounce the dagesh hazaq in tefillah or qeri'ah, does it make sense in real life to pronounce these bg"d kf"t letters as a fricative followed by a stop? Most likely, the person doing this would get corrected. This would be awkward at best. Let's say that during Untaneh Tokef the hazzan sings something about "yom tzom kifpur." A few possibilities arise. Maybe someone could incorrectly "correct" the hazzan. If the hazzan insists on his correctness, it would probably be embarrassing for the correcter, the hazzan, or both. Not to mention disruptive. But if the hazzan accepts an incorrect correction, that's not really great either. Even if nobody says anything, the pronunciation would be so unusual as to be a distraction for everyone.

It seems to me that when we have a dagesh hazaq in a bg"d kf"t letter, the best thing is to sacrifice a possible correctness for decorum. But then, who cares how it seems to me?