An estimate of 35% and a calculation of 29% is not a difference of any significance and probably accurate. The EF is not a static parameter and varies from heart cycle to heart cycle, and that with a margin of error of 10% provides acceptable difference.
My experience with EF relates to an echo test and an angiogram. The results were 29% with echo and hours later with an angiogram 13%. I brought the matter to the attention of a tech, and it was stated 13% is incorrect as if I had 13% EF I wouldn't even be able to walk up a flight of stairs without serious shortness of breath.
There is fractional shortening with heart chamber measurements that involves diastole (filling) and systole (pumping phase) parameters that may be better. For insight, the wider the heart chamber during diastole more blood can be pumped, and the size of heart chamber after systole indicates the strength of contractions.