There are definitely situations where the perceived level will be louder between peaks being limited to -1dB or -0.1 db. It depends on how hard the music is hitting these levels and how often the music goes up there.
 
If the music is only hitting a peak level say 50 % of the time then yes you won't notice much of any difference between -1 and -0.1 dB. But there are other circumstances when there is a difference. Recently I mastered a hip hop album and these guys like it seriously loud. They like it literally slamming. I don't and I don't like mastering this hard but sometimes you just have to do it. At least PSP Xenon can handle it and keep the mix relatively together.
 
At first I set my upper limit to -1dB and the music was hitting that like all the time. Then I increased that to -0.1dB just to see if there was a difference and it was definitely louder and it was quite audible in fact. It is literally a dB higher so when the music is slamming that hard yes it makes a difference.
 
A lot of people around here are not mastering anywhere that hard so it is easy to make assertions that there is no difference between -1 db and -0.1 dB.
 
I have never had a master rejected either by a CD manufacturer when my masters are hitting -0.1 db but after reading Daniels's post I think it makes sense and I might pull that back now to -0.2 dB instead (on those very loud masters that is) Luckily most of my clients are not hitting that hard and I don't have to go there that often. I agree that limiting to -1dB is good for MP3 conversion. I actually do two masters one at -0.1db and a second at -1 dB for that purpose.
 
 
 

http://www.dr.loudness-war.info/
 
 
read it and weep.
 

In PSP Xenon, you need to increase input gain to reach the threshold.   The output knob decreases the overall level so there is still headroom to protect you from analog clipping.  At -1db the output will be .9dB quieter than if it was set to -.1dB.  It does nothing to the actual peaks.  That could be a noticeable difference in RMS levels for sure.  For the same levels you would have to increase the input gain by the same amount and sacrifice dynamics even more.  
 
Some limiters will allow you to set the clipping threshold without changing the level with either makeup gain or output attenuation.  In that case the RMS levels would be the same at -.1 or -1dB.   I all really depends on the controls in the limiter.