I agree with Scott.
At the low volume parts more support would help to bolster the notes that are a little flat.
When the stronger part comes, the volume is a bit too much. Again, support can act as a shock absorber for this and allow you to put your strength in pushing down in your abdominal area rather than punching out too much air and volume from your lungs.
Look at the waveform image on the Soundcloud file. The amplitude takes a big jump when you go to the high notes. Yes, I know there is an increase in intensity at that part of the song, but it needs to be more subtle. It should be more in the sound of your voice, but not the volume. So you need this sound, but a smaller, more controlled version of it. Try really pushing down there and downsizing the volume jump. It will really make for more consistency in your tone and singing in general.
The main benefit I am looking for in your most recent example is more consistency in your voice from start to finish within that song. It produces ear fatigue to the listener if it starts out at one level and then starts banging on your eardrum later, especially if it is a worship song.
The overall end result of any vocal performance hopefully is satisfaction in the ear of the beholder.
It doesn't necessarily have to be "soft" but you don't want too much dynamic range, either.
Certain ranges of singing do take more effort to sing lightly. Some upper midrange is hard to do without notching up the volume or going into a lighter mechanism. When that happens, you still have to be careful that the volume level of that section of the song is at an appropriate amount for the feel or mood of the song.
It is a common habit for singers to start belting when something exceeds a given pitch. Higher pitches can be perceived as more piercing sounds, and may be heard as less pleasing sounds.
Always Listen to the playback and give yourself honest feedback.