Extract Recipe and Late Additions and Calculated IBU

[AHarper]

I was creating an Extract recipe using DME and was wondering if various levels of Late Additions would make any difference to the resultant IBU. What triggered my interest was the varying IBU values as I changed my mind on the amounts of DME I was going to use so I constructed a test recipe to plot out the differences.

The recipe was to be simple with a fixed Batch size with a single amount of DME and a single Hop. There was to be no Additions to the recipe and a standard yeast of an average Attenuation level. There was to be no Steeping grains and the DME split between an Initial amount and a Late Addition amount, both totalling a set volume of DME to create a standard beer. The variation in amounts was to be in 10% steps.
What I came up with is detailed below but the interesting thing -to me anyway - was the shape of the resultant graph. I have no fixed idea as to whether the data is surprising to those experienced brewers out there, and I would welcome comment here, but it does show the variance in the strength of IBU Isomerisation with percentage of DME added.

The initial value of 50g was there to provide some early stage wort to have the hops work with - otherwise it would just be water - but then the resultant IBU is calculated as LOWER than the next amount (10% DME).

There is no mention of what the Late Addition Time is in the Recipe Builder. Does this matter? Would the IBU values change with a known fixed timing for the Late Addition? Probably but I do not know how the Builder handles this. This was why I did this experiment - to see if there was a difference and effect.
As well as playing about with the recipe Builder I also plugged in the appropriate values into the Batch Statistics and IBU Calculators just to confirm calculated values - the results are detailed next to the graph below.
Comments?

The Recipe:

The Graph of Results.

 

[Yooper]

It's be traditionally said that the higher the SG of the boil, the lower the hops utilization. Later on, Palmer said he thought it was more related to break material in the wort falling out than the actual SG of the boil, but that's the same model that's been used for IBU calculations forever.

In my own experiences years ago, I did notice that adding the bulk of the extract late did provide more bitterness to the beer. I also noticed that the larger the boil volume, the more perceived bitterness as well. That has to do with the maximum isomerization of the bitterning compounds in the wort, before the wort becomes saturated. That maximum has been said to be 80-100 IBUs.