I'm not familiar with this method, but how was the calibration curve constructed? For example, does "1:1 = 1478.2 ug/mL" mean that this sample has an absorbance corresponding to a standard of 1478.2 ug/mL prepared in the same way, i.e. 100 uL of standard of this concentration plus 1.5 mL piers reagent? If so, that is the concentration of the 100 uL sample, not the 1.6 mL solution. If you multiply your found concentrations by 0.1 mL, they seem to give amounts more in line with your expectations.
By the way, you need to sort out your dilution terminology. I assume, from the ratios of the concentrations, that your samples were:
Undiluted, i.e. 100 uL stock, no added water - that's 1:0 sample:diluent.
5-fold diluted, i.e. 20 uL stock, 80 uL water - that's 1:4
10-fold diluted, i.e. 10 uL stock, 90 uL water - that's 1:9
Adding 1 part sample to n-1 parts diluent (1:n-1) gives you an n-fold dilution. Calling the undiluted sample (if that's what it was) 1:1 is particularly misleading.