From what I can tell, the chart is fairly accurate to +/- 3-6 months, but using it to say your helmet is an August 1943 McCord is fanciful. As Marc points out in the article, pressing the helmet shells didn't necessarily follow the sequential order that the lot numbers were assigned, so lot 619 could have been pressed before or after lot 620. This means there is potential for relatively low lot numbers to show characteristics of later M1s if a pallet of 1942 blanks way in the back corner didn't get pressed until 1944. In addition, after a fairly short ramping-up period, the McCord chart smooths the data by assuming a steady rate of production, when in fact the number of helmets produced in any given month varied widely. Below is a graph of McCord's actual 1944 M1 helmet totals based on Ordnance Department reporting. It's easy to see how assuming steady production could result in inaccurate dates.
Until someone locates additional information about what lots were pressed at a given time, it's best to use the chart as a general guide and not as the final word.