Line level crossover calculations help

1. Why passive?
2. Why line level?
1. When there is no room in the amp's design. I don't add stages to an amp just for frequency selection where I can manipulate the response over a stage's load or an interstage load. Here 'active' means within a dedicated feedback loop, (ie even passive frequency selection components are surrounded by active stages) and maybe the amp doesn't have a suitable loop.

It is said that passive line level delves into the noise floor. This is not a problem, as I see it the noise floor doesn't change provided there are no passband losses.

2. Line level reduces unneeded portions of the band from amp stages. Bass amps can be kept simple and components with marginal bandwidths can be worked in. Treble stages can be low swing. Gain can be designed to suit the speakers. It also gives easy loads to work with and opportunities to use esoteric small value capacitors.

why not speaker level?
The resonance impedance peak of a horn midrange can be difficult to track as it may have a high Q factor making line level a more reliable alternative. Even in general, sharing between line and speaker level can ease the adjustment process IME. The last dB or two of efficiency will be cheaper to chase down by low passing a woofer at line level.
 
I use a passive second order crossover at 300 hz.
I use 10h chokes for mine for the most part. Mine also has some LF boost, so I use about 30h on the low side.

Small signal chokes of 1Henry+ were seen in graphic equalizers before they were just loaded with lots of op amps. They were plentiful before other options existed.

The 10h chokes measure about 1x2x1.5 inches so not large at all.