It can be tricky, but my understanding is that no other voice moves off of it's appointed reciting tone until the voice carrying the original psalm tone melody (usually the tenor in a 4-voice or baritone in a 5-voice setting) does so. Similarly, no other voice sings any word of text until that word is sung first by the voice carrying the psalm tone melody. They can, however, all sing words simultaneously. Where it gets complex is during the mediants and finals, especially the finals that can become quite rhythmically complex in the other voices.

Start by pointing the text sung by the voice carrying the psalm tone (or its approximation. Sometimes the tone has been simplified.) Then match that pointing to the next voice with the least amount of rhythmic activity, proceeding through each voice part from least amount to most. It is in the heavily syncopated parts that the text underlay can become complex. You may find that it's necessary to "modify" the rhythm in these voices (such as dividing a half note into two quarters) in order to observe the rules.

Since there are few examples of a falso completely written out, this is the best information on performance practice we have, so far as I've been told by a musicologist who has studied the genre. Falsobordone were sung frequently and the professional singers of the day were familiar with the rules of performance practice.

I have been trying to determine a simple, standardized, language-independent method to underlay text for Giammateo Asola's thirty-six CATB falsobordoni [PDF], and I have made the following conjectures/determinations with regard to this collection after transcribing it to modern notation:

1. If these falsobordoni (like others) weren't published with underlaid text, then singers may have sung the psalms from a text-only psalter (having more or less memorized the falsobordoni used for a particular day's office).
   
   
2. For absolute ease of use, the text pointing would have needed to be compatible with all examples in this collection.
   
   
3. The chordal examples, e.g., del 1o tono, 3o ordine, indicate a four-syllable cadence at both mediant and final, i.e.,
   
   reciting tone ' 1 2 3 ' 4 * reciting tone ' 1 2 3 ' 4
   
   These chordal examples suggest strongly to me that if ease of use were historically a consideration:
   
   
   1. the same four-syllable cadence would have been applied to the polyphonic examples;
      
      
   2. the cadences would have been accent-independent (this opens up a world of possibilities with regard to vernacular adaptation);
      
      
   3. singers either may have pointed the psalm texts from which they sang, or may have known the psalms well enough to sing from unpointed editions.
   
   
4. To use the more polyphonic examples, singers may have determined beforehand—or on the fly!—the note(s) in their part where the first, second, and third syllables of the cadences were to be sung—with the fourth syllable always sung on the final note.
   
   
5. Nowadays, in most places, text underlay is mandatory for successful occasional use of these falsobordoni.

The attached example illustrates how they might be underlaid for use today. Ambitious choirs might appropriate it for use on Corpus Christi this year.

Interesting! If I were singing on the fly, my accent-dependent first instincts would lead me to something a little different: