There are a number of sources that confirm blood pressure cuff size can have a significant impact upon the measured reading. (See 1, 2, 3, 4) The most succinct description I have seen is this:
The error that is introduced by too small a blood pressure cuff is a reading that is too high and the error can be 50 mm and this is not rare. ...
Too large a cuff introduces the opposite error, an artificially low reading but not nearly the magnitude of the too small cuff error.
Disclaimer, I haven't reviewed the material too closely on the site I quoted, so it's likely they're trying to sell something. But their summary is in alignment with other non-commercial sites I've found.
What I'd like to understand is why the blood pressure cuff size affects the measured reading.
From my simplistic understanding, a blood pressure cuff squeezes against the brachial artery.
The compression of the brachial artery temporarily stops blood flow and allows the practitioner to measure when flow commences again.
Assuming the cuff is large enough to stop blood flow, I would like to understand the physiological principles that are affecting the read measurement. And I think this is a valid engineering question as I'm interested in the interaction between the measuring device (blood pressure cuff) and the system being measured (blood pressure via the brachial artery). In other (much bigger) words, what biomedical engineering principle is at work to cause the measured blood pressure to differ?