One of the reasons I avoid online hobbyist/enthusiast "tech" forums is they tend to prefer never-ending debate of vague, unproven or unprovable beliefs rather than take an engineering approach and look at the actual numbers to resolve the issue. As I showed in this video, we did a simple test to show the resistance of a 0.5 amp fuse is only around 0.5 ohms, and higher rated fuses can be much less. And if anyone is hesitant to do a simple test, freely available manufacturers' fuse datasheets will show the resistance of all fuses. I'm guessing that nobody thought to hook up an variable rheostat (I show a 0-5 ohm rheostat in my "What is a Rheostat" video), and match it to the fuse resistance and monitor the audio output on a scope or spectrum analyzer to see if varying the fuse resistance matters. And I'm guessing nobody thought to consider the fact that the 120v power transformer feeding the audio equipment and protected by the fuse probably has a winding resistance 20x that of the fuse (as I showed in my "Understanding Transformers" series), making the fuse resistance irrelevant. Not to mention the other circuit components, whose resistance can vary with component/manufacturing tolerances, temperature, etc. So when they say "I notice a difference in the sound", the engineering response is "show me the ACTUAL variation in audio output signal as you vary the fuse resistance".

Yes, Fuse is bidirectional.

No, they can’t. To determine the power flow direction, a device needs to have a reference, for example, voltage. With no reference, fuse will operate with no respect to current flow.

not without damaging the fuse