There are a few causes of delay for input monitoring in any DAW.
1. The audio buffer size itself.
2. Plugin's that require lookahead in the signal path. By default all audio is delay compensated and this will cause input to be delayed.
3. The input/output (hardware) latency of the audio device itself
 
If you want to track at the lowest latency you must set 1 to be as small as it can be without causing dropouts. If you still hear latency after doing this its likely caused by 2. To overcome this you can bypass any plugins that cause delay in your signal path. Alternatively you can press the button labelled "PDC" on the mix toolbar. This will ignore any host delay compensation for "live" input monitoring.

I have a dedicated monitor mixer with effects for tracking. But a long time ago before I had that system I had a client ask for reverb and quickly figured out a simple solution.  
 
You will still use direct monitoring at your interface to keep the dry signal loud. 
Turn on input echo for the vocal track being recorded. 
Have the vocal track set up with a send to an effects buss with the reverb. crank it up.  
Set the send to PRE fader. 
Now turn down the actual track level so you wont hear the delay coming out the back end. You will now hear the vocals both dry via direct monitoring and wet via the reverb buss. The reverb will include the RTL delay but because your mostly hearing the dry signal it's not a big deal. 
 
Another way is to send the Vocals to a Vocal Buss and just mute that buss. You'll still hear the signal from the effects buss. 

Few use this feature now I would think as most audio Interfaces have direct monitoring with no latency. 

From OP question, it is unclear how advanced his setup really is... Mentioned DAWs are Studio One and Mixcraft. Well, recent versions of Studio One has "low latency mode" (== in advance processing), but does Mixcraft has that?
 
Can the driver mode be the reason? Is Sonar set to work with ASIO (or at least WASAPI Exclusive)?
Have you tried to set the same buffer settings (as in other DAWs) in Sonar and check without any plug-ins in the project there is no delay?
I hope the interface is at least of Zoom UAC class (RME/MOTU/Presonus Thunderbolt), the chance to get vocal monitored throw DAW without noticeable delay is almost zero otherwise (independent from the DAW).

I always use software input monitoring with effects. I can't even remember the last time I used hardware monitoring (probably back in the '90s). I can definitely see using hardware if you're recording multiple inputs from a band or something, or if you never want to bounce, but for me software monitoring has always been very acceptable, and I'm still using a Core2 and a discontinued audio interface (granted, with very good drivers). I will say that Studio One's dropout protection using separate streams for input and output is a pretty brilliant idea and works really well, so that would be an area Cake should explore. There's been plenty of feature "borrowing" from everyone over the years.

stxx
It has real time monitoring.   Read the manual. However, its best to monitor through your interface mixer software (unless you're running thunderbolt) and have the 2 bus  from Sonar play through the monitor inputs.  This way you mix your previously recorded tracks with the new input with little delay.  Best to have your previous tracks bounced to a rough mix onto a stereo track as well so you can play to a track with some FX.   Adding plugins causes latency so I run with the FX bypassed and a rough mixed tracs as I just described but there are many ways to do it.   

I don't have clients anymore (because I just do my own stuff) but when I did I could pretty much guarantee that if I tried to feed the singer a prerecorded stereo mix she would want to hear "a little more guitar" or "a little less hi hat," or "more echo on the horns." So I didn't do it.

fitzj
Few use this feature now I would think as most audio Interfaces have direct monitoring with no latency. 

Anyone tracking with Amp Sims would probably disagree.

Agree with Larry, I never had a problem using a full project when recording vocals. Never had an issue with having the "band" fully fleshed out and basic effects on everything. There's really just a few plug ins that cause problems and I by pass those in the bin. 
 
Obviously Mr Step D is only recording VST instruments and not tracking guitar, bass or vocals live.  
Either that or He/She is oblivious to timing. 
It takes a pretty good modern computer and a top of the line interface like RME to have usable RTL that you might not notice. My system gives me about 7 ms at 128 buffer and even that is enough to be noticeable. My input echo is off unless it's midi. 

sock monkey
 
Obviously Mr Step D is only recording VST instruments and not tracking guitar, bass or vocals live.  
Either that or He/She is oblivious to timing. 
 

Yeah, I have no idea what I'm doing. I feel so silly.

Real time monitoring is best done on stage. 😄