First, my best chain: Slate or Uad tape through la2a and then api 2500. (each with 1-2 db reduction max) then the pultec, boosting 12 or 16 khz, bandwidth set to broader. After that, Sonnox eq with the low pass filter (6db/octave) set at around 17 khz. Sometimes a desser too....normal things.
Secondly: sending it to a stereo channel on which I insert
a Fairchild on lat/vert mode and sidechain unlink. 1 db of reduction and suddenly, your voice is still in the center but more wider in the stereo field. Similar to MS processing.
Third: reverb channel. I prefer two channels with each a mono reverb. Same room, but two different settings (especially for the predelay). More realistic. Tape plug before the reverb to chop off the transients a bit. Then, after the reverb, a compressor with a somewhat fast attack to make the tail bigger. Then, eq'ing of the reverb.
Sometimes too, phase rotation to adress the cancellation caused by the reverb, so it blend better with the dry track. A lot of times, doing this, my verb channel was better. More body, fuller sound.
final idea: Cool trick. Instead of using the predelay of the verb., I may keep it at 0 and insert a tape delay (with some LFO) before. Depending on the song, with the good setting, it sound magical and allow you to play the verb channel more subtly, as the ear notice instantly a track with a modulation effect on it.
These are some ideas that makes me got some superb vocal sounds. Of course, not necessarily all of them in the same song!