I have fielded a variety of requests to alter Heroes By Trade to make the strategy more complex and add elements of realism over the time I have been building it. Most of the recent requests have come from Naked Man and they have been all over the place including things like poison tables, facing, flanking, zones of control, imprecise locations of AOE attacks, support for hex grids, and more. The other thing Naked Man has advocated for is faster combats that are resolved in much less real time. Unsurprisingly these two things are very much at odds with one another. Many of his suggestions like support for hex grids have been spot on and have already been implemented but the others aren't so workable.
Strangely adding in support for hex grids instead of the default square grid setup was really easy. I had figured that a tactical game would be extremely different on a different grid design but all I had to do was design two sets of AOE effect shapes that could be usefully described with a single set of words. I eventually settled on the following:
Area 2: A 2x2 set of squares or a set of three hexes in a triangle.
Area 3: A 3x3 set of squares or a circle of hexes three hexes across.
Area 5: A 5x5 set of squares or a circle of hexes five hexes across.
Area 7: A 7x7 set of squares or a circle of hexes seven hexes across.
Simple enough, and now the game can be played on either a square grid or a hex grid with no change needed in the text. It isn't a perfect translation but it does allow those purists who hate the fact that diagonal movement is wacky on a square grid to play on hexes and have slightly less mathematical inconsistency. Slightly.
The rest of the requests have been trickier. The trouble with adding in facing is that I don't feel that it adds anything to realism. People standing, frozen in time, waiting while somebody else runs around behind them to stab them in the back isn't realistic any more than not having facing is. The important question to my mind is whether or not facing adds interesting strategy to the game without making everything take longer. I don't think there is any way to avoid increasing combat duration because if attacks are more successful from behind a target then every attack must be calculated with an additional input. One of the core principles of HBT design is that figuring which action is the best should be complicated and very interesting but once you decide what to do it should be very quick to resolve. Adding in facing definitely increases the complexity of the choice process (good!) and also increases the time to resolve any given action (bad!).
Imprecise locations of AOE attacks have similar issues. AOE attacks being difficult to aim sounds pretty cool but it would be a brutal slog to play with. So, you tried to aim at three enemies and just barely miss three allies? Time to make a whole new extra roll for every one of those targets to figure out if they will actually be affected. Most likely what this means is that AOE effects simply aren't usable except against groups of enemies that are far away and while it strikes me as reasonable that you don't set off huge explosions near your friends it certainly increases resolution time while adding little to strategy. I have definitely played with players who would always cast Fireball and end up hitting their friends and players that would never cast Fireball if there was even a chance of hitting a friend and both tended to frustrate people to no end.
Zones of control and flanking were standard in DnD 3.x and while they did have their advantages I think they created more problems than they solved. Flanking is a pain in the butt to define and explain to new players and has the same problems as facing in terms of constantly altering the math for every attack. I can't get behind a mechanic that means that melee characters are bad without a teammate and which slows resolution down tremendously. Zones of control on the other hand have a huge variety of implementations, some of which are good and some of which are a disaster. Currently HBT uses zones of control but the only thing they do is penalize ranged attacks. Using a DnD style mechanic where moving past a creature entitles it to a free attack pretty much means you either have a way to negate that attack or you simply never move past a creature. If I could find a very easy to resolve mechanic that didn't come down to 'never provoke an attack of opportunity ever' I would be interested in using it but so far nothing has come to mind.
Poison tables are actually a thing that doesn't have this sort of tradeoff. You can have a big scary table with tons of poisons on it that is rarely used and that doesn't hurt anyone else at all. As long as a feature doesn't bog down the standard game I don't have any issue with it and making a list of poisons and the ways in which they can be stopped is not the sort of thing I was really aiming for the HBT but it is entirely harmless.
Feature creep is the thing I struggle most with when it comes to player feedback on my games whether they be RPGs, board games, calculation tools, or computer games. The great majority of feedback comes down to 'add this list of extra features' that all sound great but which extend the playtime of the game substantially. (The second most common request is 'make the game play faster'.) I actually design my games to a particular timeframe and that means constantly rejecting new feature requests on that basis. Not that I object to new feature requests - many features like hex grid support can be added with no downside. Telling people that the thing they want is never going to happen is tough, especially when it sounds cool to me too. I have to keep my focus though as there is no shortage of very deep, incredibly complex games that absolutely nobody plays because they are brutal to get into and take forever to play. The world has more of that then it can ever use.