This isn't correct. Per the IEEE papers, the inaccuracy is due to high levels of conducted electromagnetic interference from switching devices, not the wave shape. Meters with Rogowski coils to measure current have positive errors (the customer pays more), those will Hall effect measurements have negative errors (customer pays less) and those with current transformers can be either but of a small amount. For residential customers this could occur for those with a lot of noisy switching on their lines, like grid tie solar panels with inverters. I'm sure the utilities are working with the meter suppliers to correct the problems. They are required by the state regulator (and commercial law) to have accurate measurements of what they sell. Food for thought: How do you know your old electro-mechanical meter was working accurately?