Ofcourse, in a 24dB/oct LR active XO, exact component values are essential so that the high pass and low pass matches each other good to get a smooth transition between the drivers. Most advices I've seen is to use close tolerance (1%) KP caps and metal film resistors.. But what I cannot figure out is why people would put a resistor (or ten) consisting of feets or meters of metal film in the signal path? I do not have alot of experience in this field, but I *will not* spend top $$$ on signal cables knowing that the same signal would pass through extensive lengths of ultrathin metal film 
It would make me feel stupid, as I could just as well roll the film out, attach couplings, and use it as the worlds worst signal cable...
I'm planning on a simple 2-way crossover, my first one, with 4 gainclones, to feed my bookshelf speakers. I will avoid the metal film in signal paths, using instead 1% Riken .5W carbon composite resistors, wich I will order in 3 times the required quantity, using only the best matched 1/6 in each crossover channel. For the "double" resistor to ground in the high pass circuit, I intend to use 0.1% Welwyn metal film.
For capacitors, I can get reasonably priced 1% Silver mica caps, wich seems ideal, or are they?
😕
Decoupling of op-amps will be done with 0.01uF ceramic (maybe with open slots for adding an extra 0.1uF ceramic) as these have a great ability to kill HF noise on the +/-12V supply lines.
Other circuitry will use Riken resistors in signal paths, Vishay and Caddock resistors, PS will use Black Gates and Phillips LongLife caps. Input coupling cap will be 1uF Hovland Musicap, and for information, I will start off with samples from TI; OPA2134 op-amps, but ofcourse with sockets so I can replace them if desired...
Does it sound like a sane approach? Have I missed something?
It would make me feel stupid, as I could just as well roll the film out, attach couplings, and use it as the worlds worst signal cable...I'm planning on a simple 2-way crossover, my first one, with 4 gainclones, to feed my bookshelf speakers. I will avoid the metal film in signal paths, using instead 1% Riken .5W carbon composite resistors, wich I will order in 3 times the required quantity, using only the best matched 1/6 in each crossover channel. For the "double" resistor to ground in the high pass circuit, I intend to use 0.1% Welwyn metal film.
For capacitors, I can get reasonably priced 1% Silver mica caps, wich seems ideal, or are they?
😕
Decoupling of op-amps will be done with 0.01uF ceramic (maybe with open slots for adding an extra 0.1uF ceramic) as these have a great ability to kill HF noise on the +/-12V supply lines.
Other circuitry will use Riken resistors in signal paths, Vishay and Caddock resistors, PS will use Black Gates and Phillips LongLife caps. Input coupling cap will be 1uF Hovland Musicap, and for information, I will start off with samples from TI; OPA2134 op-amps, but ofcourse with sockets so I can replace them if desired...
Does it sound like a sane approach? Have I missed something?