Is there a specific reason why you cannot reduce the particle size any further? Perhaps, more information about the problem and your trials could help you attract better answers.
dissolution does not matter for protein considering you anyway have to carry out digestion per Kjeldhal method. If it is too coarse, only then the moisture and fat estimation could become error prone. Therefore, it would be ideal to perform a fine grinding before analysis; or sieve a large batch of the flour and collect the finest fraction.