What I found in my own game, which was rather SuperRuneQuesty ( 😜 ), is that opposite issues could emerge both with creatures larger than average and those smaller than average -- and in the latter case, those could easily involve player characters if people wanted to play a duck or a value trollkin or etc. -- and that was on a RQ3 rules base.
But the counter-intuitive extra vulnerability of creatures larger than average also affected my ability to let players run say a dark troll or a morokanth ; while conversely the extra damage that they did from their extra SIZ+STR was also imbalancing.
And I came to the conclusion that BRP/d100 works well where every player character is human size or equivalent, without exceptions, but the ability in a Gloranthan or some other fantasy or science fiction setting etc to play a character a lot smaller or larger than a human leads to having to deal with these things in one's gaming sessions. (it worked really well in the Stormbringer game)
It's something that I never quite succeeded in finding a completely balanced and easily applicable solution for, though I did manage to come up with some design principles that seemed to help with some of the basic problems, that could even be viewed as rules bugs.
1) Shift all base human base stats up to 2D6+6 -- and then scale around average stats = 13 instead of 10-11
This one involved a LOT of detail work that I never came close to completing, as it means you need to re-stat every creature in the Bestiary ; but I was doing it piecemeal and as needed for my immediate game needs
And this change is mainly to make the ability to create characters smaller than average a bit easier, by increasing their potential range of stats into something more workable ; it also potentially opened the possibility of such things as timinit player characters and so on -- the knock-on effect on creatures larger than average actually worked out to be rather minor, because of the way RQ scaling works
2) Change all STR+SIZ damage bonuses to positive values, instead of having negative values for creatures smaller and weaker
This is easy in principle to implement, BUT the knock-on effects into the combat system are extensive -- because you end up un-balancing basically everything in it
Again, I never quite managed to re-balance this, except when it only involved creatures of average human size ; but accepting that as a compromise clearly would have ended up defeating the entire purpose of my house rules changes, so it wasn't good enough
It created secondary issues of what STR+SIZ tiers to use in the damage bonus table and what level of granularity and which general scale & rate of increase, but playtesting would eventually have solved them, so they weren't systemic
3) Balancing headaches & cans of worms
I *did* manage to work out as a principle of (** cough !! **) "realism" that the larger a creature is compared to a SIZ 1-3 base line, it should have a correspondingly larger kinetic/inertia resistance to being damaged, which balance should dictate as being cancelled out in relative terms between creatures in the same SIZ bracket, but then that opened another can of worms, as everything also needed to be balanced towards a human average stat of 13 ; trying to get both of those goals to work seamlessly and in conjunction is something I never quite achieved -- including because you can only really see if something works by playtesting it, and it required a LOT more playtesting than I ended up being able to manage
And the working temporary rules that I had to emulate that kinetic/inertia resistance to being damaged idea were very clunky -- I was starting to realise that it needed to be integrated into hit points and armour points instead of being an extra and new layer of protection, but to what ratio and with which values I never managed to work out before permanently leaving Paris and sadly no longer being able to play RQ ; I was also starting to realise that I would have needed to revise the damage values of all weapons (I don't think I needed to touch the values of pieces of armour, they were self-balancing from their RQ3 ENC & weight numbers)
If I'd had access to MRQ at the time, I would almost certainly have adapted the rules principles for HP from that game, as they do make many of the scaling issues of RQ2/RQ3 and now RQG that much easier to house rule -- just getting rid of the general HP stat as MRQ did helps get rid of a lot of the imbalances that can exist in very small and very large creatures, between general HP, hit location HP, and incoming weapon damage
Another general problem that I had was in trying to make sure I wasn't designing a completely different game to RuneQuest -- which again, would have defeated the purpose. These were substantial and massive changes to basics in the game design, so I always had to keep myself reigning things back towards the RQ "zone". or at least towards a BRP/d100 HeroQuest one.
It was and to an extent still is frustrating to me that I never managed to iron out all the kinks in that house rule design ... it was particularly difficult to get the damage bonus / weapons damage / AP / HP fully balanced but still fun and recognisably RuneQuest at every possible scale in my game ; but the design that I had probably never surpassed 65% - 75% project completion