Various people have worked on (various parts of) the Estabs at various times. We've done efforts like you describe, starting from TOEs for unit estabs, substracting values (vehicles, drivers, messengers etc.) from them and adding these to their respective HQ, Base and/or support platoon(s) etc. But to work out a complete set of comprehensive conclusive rationales for this is nearly impossible given the quantity of and variation in units. And then once these Estabs are being used in a Scenario and set at a randomized unit strength percentage, decreasing both personnel and equipment accordingly, one quickly realises - however painstakenly the effort on the Estab level might have been started -, it can only be as much: without exact historic nominal unit strength data (and, except in some very rare cases, how to find these figures at a reasonable investment on company level, let alone broken down according to the same rationale as used for the game engine?) this can only be an approximisation. Next to that other factors, either from real life (such as national differences in doctrine), or in relation to the game engine (like how it handles mount- and dismount routines) add to the fact that calculating Unit Estabs and derived scenario unit strengths can never fully become a 'rocket science', although the starting point for development has always been the TOE or KStN.
Short answer for the concrete example: taking out the drivers and gunners personal weapons makes sense.