It's confusing, perhaps, but consider the following:
1) The HF curve location is a function of the HF driver impedance. Note in the second curve posted by Giskard cited above, it shifted rightward ~2 kHz as he changed the impedance from 8 Ohms nominal to 3.5 Ohms actual for 2425H. We have to measure or otherwise know the impedance of the HF driver for the curve to be accurate.
2) Where the voltage drive curves cross is not the crossover frequency. The HF curve must be raised by the difference betwen the sensitivities of the HF driver and the woofer, and their response characteristics at the frequencies of interest, to approximate the system crossover performance.
John W wanted to know if using the LF section from 4425 would yield a better result than the "standard" NL200t3 when using a 2214H woofer in his project:
http://audioheritage.org/vbulletin/s...ead.php?t=7110
Here's the two LF curves shown with a 4-Ohm HF impedance. It'll be interesting to see how his speakers actually measure: