What are some favored / subtle delay settings for vocal - say for example on the Sonitus -fx delay. There are so many parameters that I get kind of confused on how to set it up right. Thanks ! 

The best way to learn it is to try it out.
 
For delay times, here is a quick overview -
                0-15/20 ms is not distinctly discernible and manifests as a timbre change (comb filtering kicks in early and begins to drive the timbre down). This area is good to play in to understand how the sound changes with adjustment.
                20-50 ms is just discernible, and manifests as a "doubling" effect (this is where "chorusing effect" kicks in).
                50-150 ms are easily discernible and sound as rapidly decreasing echoes (or "slapback echo").
                >150 ms delivers highly defined echoes (not often useful for voice, depending on application).
 
As far as feedback, also play with this... lower values are often best with voice. The delay is a filter loop that is introducing the original signal back onto itself modified by how you adjust the settings on the delay.

No one has answered because there are no 'best' settings and there are no correct settings.
 
Its all relative to:
1.)The song,
2.)The songs tempo
3.)The vocal
4.) How you want it to sound
5.) what kind of reverb is on the vocal track ( reverbs are a series of short delays)
 
Just learn the settings on any delay as the settings are mostly universal for all delays and then you would use your new found knowledge and your ears to decide what works best for that particular vocal track and song.
 
I hardly ever set a delay the same for 2 different song, so what will work for one song, may not work best for another song. When you understand this, you are on your way to becoming a better audio engineer
 
Here are some vocal recording and mixing tips that I have written down:
http://www.audio-mastering-mixing.com/faq---q-a.html#24
 
CJ

I find it interesting that Mettelus ended at 150 ms,, my lowest setting.
Using tap tempo or the tempo mapping featuer is a good way to sync the delay to a song. 
I havn't had as good as results with the VST efxs as I get with my Yamaha 01v's delay. 
I used it live for years and it was pretty cool to be mixing a live band and tap in the tempo and then using the note values achieve a good setting within the first verse of a song.
Worked best to enhance Sax or Guitar solos. You don't want to overdue it on vocals as it makes the lyrics muddy.
 
But the most common setting was a low feedback,, just a few regeneration's,  
And 175 ms for fast songs up to 250 ms for slower songs. 

I often use 150 to 250 ms to double the voice.  I personally don't like the way feedback sounds so I never use it but to each his own.

I like ping pong delay, set to tempo of your song or sync, Left delay at the eighth note, Right delay at the quarter note, 30% feedback. For ballads I'll take the output from the delay and run it into my hall. Delay is less in your face that way.

maybe work through the presets of the VX-64 for some ideas.

I see you've gotten the typical "There are no such things as good presets" answer a few times already, and I will assume the wisdom of this answer has made a decent impact.  It really is true.  I will also assume this answer pisses you off as much as it does me. ;)  So here's my take on a practical answer.
 
I typically use delay to create space for vocals in several different ways.  But I will give you an example for use on a focal (read: lead) vocal track.  And even though you will see the setting I used - I want to spend a moment explaining WHY I used these settings so you will be able to create your own spaces based on what you want for the track.
 

 
OK.  So this is a screen grab from an actual track I did recently.
 
First of all I never use the Sonitus delay from a preset.  Never.  It is just too easy to use it intentionally for a specific purpose and get a tremendous result.  In this case I was going for exactly the thing you describe. A subtle delay effect designed to impart space.  Usually I will use timings based on the tempo of the track I am working on.  In this case I used 1/2 and 1/3 quarters on the left and right respectively.  This puts me in a fairly slow version of this sort of delay.  Using metallus' chart above, I would usually be shooting for something in the 40-120ms range, but this track sounded good with the 1/3-1/2 setup.  More often I will land in a 1/3-1/4 setup or 1/4-1/8.  Generally I will flip through some of these sorts of settings to see what sounds nice in the track.  It is important to note that you MUST use asymmetrical r/l timings to create space, although locking a (mono) single timing can also sound "right" if you want that.
 
I will usually bring feedback and crossfeed (which is just another feedback but one that blurs the image) down fairly low.  Often I will put them up a little higher than what you see here, but I am mitigating the longer delays in this case with less regen.
 
The filter setting is also important.  We just don't need to be creating any excess buildup in the <500hz range usually, and frankly it is also slightly anomalous here that I have a lower than usual  highpass @`200Hz.  It sounded good with this guys vocal.  And I usually want to cut off most of the frequencies above 5-6k especially avoiding the sibilance range for the singer.  But again this is TOTALLY taste dependent for the singer/track/production.
 
While I am tweaking these values I will have the mix percent much higher than the 6% I landed on here....  then I will adjust down to taste.  6% is in the lower half of my usual range for this purpose.  Again this most likely is chosen in part to mitigate the longer delay time.  Sometimes I am up around 9%, but rarely will I end up in double digits.  At least not for this purpose.
 
Now...  this was a lead vocal.  Had it been a BGV, I might chosen  similar settings overall, but diverged with the filter and timing, and I usually would create an asymmetrical panning for the right and left channel unlinking the r/l amounts and doing something like 7%/13% or vise versa.  I can be more generous for BGVs with just about all the settings.  I will make timings longer, filters less draconian, and regen more generous.  And especially I will push the mix slider to the right more than I will on a lead most of the time... Also in the spirit of the stereo image I'll put (much) longer/louder delays in the opposite channel from where I've panned the BGV (especially since the effect is usually pre panning)... but then I am a sucker for lush soundscapes.
 
Hope this helps.

My fav is factor 1/8 one channel and 1/4 the other, feedback 50%, no cross feed, just a tiny little bit of delay in the mix.

I appreciate these tips-- give me a lot to try. I like that these last posts are pretty specific on settings --- Should I assume that you have the delay as a send and set up as "post".