Sum is off by .01 according to calculator

This pricing calculator almost finished and as I test it there is a rounding issue somewhere that I cannot seem to find. Hoping a new set of eyes can point out something I seem to be missing. Every helper cell that is calculating is showing the correct number in that cell and in the sum cell it is calculating each helper cell correctly however the final sum is .01 off
If I round the cells the helper cells that would correct the .01 difference then the sum cell creates other test sums as .01 over.
I'm not sure where this is going wrong. Can someone please help.
Please see workbook the correct are shown as well in correct. These numbers are based on the Green totals with their calculation compared the those same equations on a calculator...

These numbers are based on the Green totals on the right.
There is one as a subtotal and the second it the sum total.
This sheet has continued to change and with my skill set I felt it was best to separate the helper cells and have them all be very granular. So from top down there are the initial calculation helper cells. From there there are additional help cells that performs sub calculations. The bottom, that has each detail as to what info is being pulled from what helper cells has those details to the right.
From these helper cells they build an equation to properly calculate the sum price.
Does that help?

I'm sorry for not being clear. Each sheet the header states if the calculations are correct for the subtotal and the sell price (green cells). If I round I21 then it raises the sheets that are correct to .01 over whereas right now where they're correct - I21 is the subtotal. For example on sheet Pricing (3) the subtotal is 7.77 and the sell price is 10.48 - that is correct
On pricing (4) the subtotal should be 79.78 and the sell price 116.28 - now here is the issue though. This workbook is actually only one sheet. It is only in book form in order to show this forum the different totals I a trying to achieve and how they're different. The first Pricing sheet is the only one that will actually be used, not the others (again only there for example purposes) Therefore if I raise the rounding cell it will change the 7.77 to 7.78 which is incorrect..
I hope that makes sense.

Is it as simple as that you want to round the final total in I26 up to the nearest 0.01? i.e.

=CEILING(SUM(I21+I8+I11+I12+I15+D11),0.01)

I'm terribly sorry for all the confusion I am removing all the extra bulk and will re-upload the sheet to help eliminate all the extra nonsense.

After reviewing with my coworker - we believe the math that was used to create our reference sheet is using rounding in some equations and not in others. I believe the reference sheet (on paper) is utilizing financial amounts in mind, For instance, 0.125 but 0.13 is used instead. Division quotients are the real "culprit" but they are technically more correct. This would still be inconsistent due to the nature of the numerical manipulation. In my opinion this would be counter-productive for the end goal which is to accurately calculate mathematical computations for measurements needed. Waste is already accounted for in the calculations so there is no apparent need for any less precise calculation. I greatly appreciate the time spent thus far on this and once I have a greater understanding for which cells "require" to be altered and if problems arise I will be sure to ask for help.
I have updated the sheet if anyone would like to see with the above corrections as well as reverting it back to 1 sheet. If there are any suggestions you see fit please let me know, otherwise, I will update the post when the final corrections are made.

Originally Posted by Cnuessle

. . . rounding in some equations and not in others. . . . end goal which is to accurately calculate mathematical computations . . .

Consistency is necessary. Usually best to perform ALL calculations without any rounding and just DISPLAY then rounded in summary exhibits. Better to adjust the least important constituent value than bottom-line rounded results. If a sum of rounded addends were 0.01 off from the rounded sum of the unrounded addends, add or subtract that 0.01 from either the largest addend in absolute value (simpler) or the addend with the greatest difference between rounded and unrounded value with the appropriate sign.

Cross-footing can also be a problem, and 2D makes it more difficult to adjust rounded addends.