The "bypass the big electrolytic with a small film cap" craze began in a time when most electrolytics dead sucked and there was no alternatives available so it would be comon to see multi-capacitors made up of a single BIG electrolytic,a poly film cap as big as you could get whih usually meant 10uF for the "mid rangew" and finally a tiny film cap of arounf 1uf for the "treble".
It was an option many follwed because the sound was totally different and the conclusion was
better and this being DIY audio once a thing is proclaimed better it is "lemmings to the cliff" time.Everyone follows and never questions or does their own thinking (including me once upon a time
)
The basis for this "improvement" was that a large cap is slower than a smaller cap and music is fast and the higher you go in frequency the faster it is so it just MUST be better to use a cap to match the speed of the music in that band of frequencies !
Then there is the other school of thought and the one I follow.One that my ears tells me is the better way.
It came to me after reading about loudspeakers and time delay distortion that a damn lot of effort was going into making each drivers sound reach the human ears at the same time in order to maintain "time" integrity.that is where the upper end sound of the woofer does not come in a tick later than the lower edge of the midrange or the upper edge of the midrange arriving after the lower sounds of the tweeter because this "disjoints" the music instead of presenting it as a single sound as it was recording.
I am reading this and the "hit with a hammer between the eyes" light bul goes off !
while the loudspeaker industry is trying to correct a known fault in systems design we DIY knuckleheads are introducing the very same thing into out electronics with these capacitor "crossovers" where each band of signals is travelling through a capacitor at a different speed depending on its size then mixed togather as a disjointed out of alignment mismash of sounds !
The tiny cap being the fastest means that is the first sound through with the 10uf next and finally the big electrolytic and while we are talking about uS (microseconds) it is heard as a blurring of the notes when the same note is added together (the overlap region of the cap crossover) at the differnt speeds of the caps.
My attitude is better a slight roloff off frequencies that are single cohesive signal than all the notes but off-timing.Damn hard to pick out individual instuments in a busy passage and actually follow what the artist is playing when that note is broken up into three segments and blurred at the crosses.
single cap of the highest quality you can afford at each position is in my opinion the only way to fly.
YMMV