TomBrooz, I have been struggling with the same questions you have since I decided to change up my extract additions from 60 minutes to at flameout. Just that change alone ended up changing my estimated IBUs from 31 to 63 (from sort of hoppy to way hoppy). Right now I am using Brewtarget to enter my recipes and it uses Tinseth's calculations to get IBUs. I am sure other people on this forum are using Rager's calculations to derive their IBUs and I am sure if I used it for my recipe I would end up with something different.
With that being said I have been searching for more definitive answers about utilization and boil gravity just to get things straight in my mind. As Yooper said in my own post the theory regarding utilization has changed. It seems the theory now is that utilization is not clearly dependent on boil gravity in of itself, since we typically think of higher gravity wort = more sugar and lower gravity wort = less sugar (there still is a correlation of sorts though). I was listening to a podcast by Jamil Zainasheff and John Palmer from last year and they were discussing this very thing. The new theory is that with higher gravity worts (see it still has something to do with gravity, but not for the reasons we usually think - i.e. sugar content) you are providing more proteins, hot break, and cold break materials that the slightly soluble alpha acids in your wort will cling to and get taken out of solution and be removed with those waste materials.
In their discussion they did impart some information that was more useful to me than helping me to understand what the new theory is. They recommended to pick a model (whether it be Tinseth or Rager or some other) and stick with it for your own brewing process - don't jump around and just pick a model because it is telling you what you want on a given brew day. Basically they seemed to think these models should be learning based in that as you used the calculations you will learn what 35 IBUs are for your recipes/brewing process. If you taste your beer and you have a calculated 35 IBU and you don't think it is hoppy enough, kick it up to 45 the next time you brew and see what that gives you. Ultimately you will learn what 15 IBU vs. 30 IBU vs. 65 IBU means for you as a brewer and that will give you consistency. If you jump around with models you aren't going to have any consistency. Makes sense to me.
Ultimately the most reliable calculator is the one you use consistently. Hope that helps you with your journey. I for one am just going to take a leap and stick with Tinseth and adjust my recipes from there.