The so-called high precision multispectral camera manufacturers argue that a more precise sensor, narrowband filters with an incident light sensor can provide valuable photometric measurements for NDVI calculations. It is unquestionable that a modified consumer camera is not good for photometric measurements. I think this is not an objective of a farmer for vegetation mapping and not even for NDVI analysis. The objective of the farmers is to have information on their fields for the lowest cost without technical understanding.
The advantages of precision cameras and their photometric datasets are valid in lab environment. The photometric reflectance of a field is affected by many environmental factors (position of the sun -> shadows; strength or direction of the wind -> position of the leaves) which can not be taken into account and which make the precise photometric values almost useless. If we want to compare 2 precision imagesets of the same field made in different time, the results won't be as precise as it is said in the promotions due to the affecting factors. The results are almost as relative as if the images would be made by a converted consumer camera. But the latter one is much cheaper and usually have a much higher resolution.
You can try it for yourselves with any NGB (NIR+green+blue) imagesets here for free:
NDVI Camera | DJI NDVI upgrade | AgroCam | NDVI image processing
An NDVI analysis made by a converted camera can also provide valuable information on the vegetation status of a field.