*** UPDATE ***
I finally hunted down a great thread that also has references within it to a few more good discussion.  So just wanted to share this in case others have the same question I did.  
 
http://forum.cakewalk.com/Lets-all-TEST-our-Interface-driver-for-offset-m3178396.aspx
 
I'm rather leery of messing with a stable system but may attempt this test at some point.  
**************
Hi Guys,
 
Please help me understand this.....

I read just recently on this forum that allegedly there is some kind of latency/time-alignment issue with WDM drivers vs. ASIO.  Having to do with the WDM drivers not reporting something properly to Sonar?  Anybody able to shed some more light on the subject?  It still is not clear to me what the potential impacts are, if any, for any given audio track.  If there are latency/alignment issues are they consistent or random?  And what kind of difference are we talking about, something entirely trivial or not?  The implication being that audio tracks recorded with WDM will often not be properly time-aligned whereas ASIO will (or at least much more precise)
 
BTW, I'm using WDM/KS RME drivers with a Fireface if that matters, and very stable.
 
Sonic

I am not able to clarify what the technical difference between the two are, just that out of personal experience that my (all up to now) Audio Interface(s) has behaved better and more stable inder ASIO.
None has behaved better than RME.
Others have to explain the technical side of it.
I ask again.. Why not try and see?
If it doesn't work for you, change back to WDM/KS.
All the best.

Only reason not to just try is simply that I've learned once you go changing drivers or interfaces it can alter settings in the system such that it can sometimes can be hard to get things back the way they were.  Things sometimes change that you may not even be aware of, in spite of trying to document things.  So the old saying "if it ain't broke don't fix it" really applies.

Sonic

Hope someone comes along that can clarify the differences.
Best of luck.

I just thought of something that seems to have been omitted from discussions on this topic....metronome.   Maybe I'm wrong, but seems to me since there is this big debate over timing/driver alignment WRT audio tracks then wouldn't the metronome itself be a potential factor as well?   Maybe in a sterile testing environment tracks can be referenced and piped around for driver alignment comparisons, but in the real world we all record to a click.  Some record the click, others use a metronome.  So how does that fit into the equation?  Especially with an audio metronome the audio is piped out of the DAW and used by the artist to react against.  Is there additional latency due to the audio metronome itself?  By how much is Sonar as opposed to D/A conversion or other factors?    
 
Just saying.   If I'm entirely wrong clue me in, but on the surface it seems like the metronome aspect would play a factor in this discussion as well....
 
Sonic

Well if you read that thread I thought that explained it all very clearly. And my screenshot say's it all.. Just look at the results between ASIO and WDM and even worse look at MME mode. Why would you want this? 
And as you see reading that thread everyone more or less had the same results using different interfaces and setups. 
Bottom line... only ASIO will sync up and in most cases within a few samples. Not enough to matter. But the amount that WDM mode is out does matter.. and the more overdubs the further your timing can drift.
 
There is absolutly no daanger in changing driver modes. Sonar will run the Wave Profiler when you switch to WDM or WASAPI. You can certainly change  back anytime and nothing will have been altered. 
I run in ASIO exclusivly but when I am recording my Band live I switch to WDM so I can add our Yamaha mixers USB giving me 2 more channels. Back at the studio I switch back to ASIO to edit and overdub. 
 
It takes only about 10 minutes to perform the loopback test. All you need is a short patch cord. 
Here once again is one of my screen shots. This one I was testing WASAPI. I also tested my little Behringer USB box. 
 

 
 

What about my observation WRT the metronome?  Have you done any testing to see how much that might factor in...?   Is the metronome perfectly & consistently aligned?   If not, that's just as much of an issue in my eyes. 
 
I'm using S5 mostly, Sonar didn't introduce proper ASIO alignment until S6.  I did buy S6 but haven't installed it, not sure if I will as it does something screwy with LiveSynthPro. Last test-install I recently did S6 completely disabled LSP, as in broke it. Even after trying to re-install LSP wouldn't it wouldn't work any longer, not just in S6 but prior versions.
 
Maybe the reason I'm not sensing a problem is because I always record to the drums one track at a time.  I never reference to anything else so I don't get the cumulative effect.  Only vocals are ever cut to the music and those are not typically an alignment-critical "instrument".
 
I am quite curious now though about just how far off the tracks might be.
 
After all this reading I've been doing there is one common element that seems to surface: The buffers sample size seems to impact things.  Lower buffers can result in less delta WRT alignment problems.  
 
Sonic
 
 

FYI.. When doing a loopback test.. Changing buffers or sample rate has no effect on the results. 
The timing offset issue is strictly a Driver comunicating with Sonar issue. Don't get RTL issues mixed up with Timming offset issues. In my thread some people thought we where testing RTL.. That's a different test and Sonar does this automaticly and seems to get this riight when compaired to using a RTL utility like this. 
http://www.oblique-audio.com/free/rtlutility
 

I cannot answer your question about the metronome as I rarley will use it prefering to use a drum track to start a song. I have a bunch of my own patterns I dumped from my Roland 505. If not I'll use the step sequencer to make a simple Kick/Snare/hi hat pattern and drag it out. I find the metronome hard to play along with. 
 
I guess you could test the metronome in a loopback test. 
Using a loop back patch cord output to input record the metronome to an audio track, Zoom in and see if it's right on the time grid. 
Because your using such old versions and whatnot there's no point any of us testing because our results will be different.
 
 
Sorry to ask,, but Why are you not willing to update? Things have improved soooo much and myself I could never go back to any older version. The midi editing alone for me has been more than the ticket price. 
If it's a money reason Home Studio is only $50 and Artist $100. That's dirt cheap for what you get. 
 
At some point you'll no longer be able to function if you stay that far in the past. And it's hard for most people here to help you because we soon forget how things worked in the older versions. So one of the perks to keeping up to date with any software is the quality of help will be vastly improved.