I’m trying to figure out whether to upgrade my Sonar Studio X2 and ditch my ProTools 8. The Gain in PT AudioSuite will analyze my recording, and adjust the gain to an ideal gain level. Does the current Sonar have a similar feature?

No. I'm curious as to how that might work.
 
Is it a feature you actually used? If so, what did it consider "ideal", how did it arrive at that number, and how does it adjust all the tracks?
 
I'm guessing it could adjust every track's trim until the full mix peaks are under some target value. But it seems that would only work if all your tracks were in the ballpark to begin with. I'm having trouble imagining how one might automate a process that's both time-consuming and different for every project. Would be cool if it worked.

After I have the tracks I want them, I run the AudioSuite Gain. It analyzes the track and gives a me a value such as -19.6dB. I then adjust the gain slider to about half that value (about +9 dB). When I click 'process' the program adjusts the gain and resizes the wave accordingly. I use this every time I record. It allows me to get as close as possible to the loudest recording without distortion. I am hoping that Sonar has something similar. I use Sonar for MIDI sequencing. But I would like to use it for recording also.

Re: the AudioSuite Gain
While it sounds interesting I can't really say that I see the benefit. 
Since the gain adjustment is done post recording it wouldn't make the signal any cleaner; your signal to noise is still going to be the same if not worse after processing.
Plus the individual track volume may have no relationship to final mix volume.  So you may be increasing the track volume on something that may have a low volume in the final mix (processing in more volume now that you will have to attenuate later).
Personally I'd rather just adjust the final gain myself during mixing.
 
Edit: recording in 24bit or higher drops the signal to noise level to the point where HOT recording levels are no longer critical. As long as the level is decent it's better to focus on the quality of the tracks and the performance than on hot levels.
You probably know all this...
 
Peace out,
...Steven

There is normalize, which doesn't give you hints about the mix levels, but you can adjust it to -9 dB or -1 db or any other level YOU choose.  Normalize every track to - 1 dB and they will all peak at the same level, as loud as can be accomplished.  It is much the same thing Gain, without the helpful hint levels from the analysis process, but will overcome any low recordings (see the Steven's post about levels - the problem is with the recording process not mixing). 
 
However, it is easier to just use faders and ears.  It does take some time to learn, but if you have recorded the tracks to put into a song, you hopefully have some idea how they work within in the song's structure and arrangement; how loud, etc.  Once you figure that part of the process out, it is pretty easy to set up levels from experience.  Personally, I'm waiting for PT Play, which chooses which notes the song plays for you. ;-)

I would think that the value of such a process would be to make sure things aren't hot, as opposed to "get as close as possible to the loudest recording without distortion". The latter is old-fashioned thinking and counter-productive in the digital world.
 
Recording hot was desirable in tape days, in order to maximize SNR. But in the digital world it's actually best practice to NOT get as loud as possible, but rather maintain at least 12 dB of headroom. You can record at much lower levels in a DAW than would be practical with an analog setup. With tape a little saturation is good. In a DAW, getting even anywhere close to clipping - even if it's not actually clipping - is undesirable prior to mastering.
 
Even if SONAR did have this feature, or if it was available as a plugin, I probably wouldn't use it. Setting levels properly calls for a human brain that can take into account musical vision, song arrangement, psychoacoustic perception, temporal and frequency masking, dynamic variation, spectral distribution and format requirements.