My first foray into bi-amping around 1987 was a total disaster: I tried bi-amping my Magnepan MG 3A using an ARC D115 mkII on the Treble/Midrange and a very powerful (400W) solid state amp on the bass that I had on loan from a dealer. The Magnepan passive XO-1 was used to split the signal. I was unable to adjust the volume level of the solid state amp so the sound was unbalanced and distorted (too much bass). Fast forward to 2005: I was now using an ARC VT100 mkII to drive my Magnepan MG 3.5. The sound was good but was limited in ultimate dynamics, bass impact and occasionally experienced clipping on crescendos. I had a pair of Kenwood L07 monobocks for several years in my closet. So I decided to biamp my Maggies using the L07s on the bass and the VT100 on the treble/midrange. However, this time, I purchased an active electronic x-over from Marchand (XM126). After much experimentation with slopes and levels the resulting sound was completely transformed. Dynamics, bass impact and control and overall sound were significantly improved.
With the experience gleaned from biamping my MG3.5, I was able to bi-amp my Magnepan 20.1 with excellent results. Bi-amping allowed me to keep my VT100 mkII to drive the treble/midrange, with the L07s driving the bass. Otherwise I would have to purchase a more powerful and expensive amp to drive the 20.1 full range.
In theory, the insertion of an active x-over devise in the signal path between the preamp and amp has deleterious effect on the transparency of the system. However, this was ameliorated to some extent by my selection of a tube x-over. Additionally, I built my own x-over modules using top grade caps and TX-2575 low noise film resistors. Thus the loss of transparency was virtually unnoticeable. In summary, biamping can be beneficial to the sound of your system, but careful matching of amplifiers is imperative and requires a quality active electronic x-over.