One issue I found 3-4 years ago was that some common tests undercount D2 by quite a bit (I forget if it was 20% just for D2 or I think it might have been 20% overall D2+D3 in some test samples). I found this because the test my doctor ordered had a statement about using a different test if supplementing with D2. A number of products (like soy milk) contain some amount of D2 so in some cases low vitamin D might be due to inaccurate testing. I haven't checked if the tests have improved in the last few years.