An interesting thread, I must say. All the way through the development of WAB one of the things we experimeted with the most during playtesting was stuff addressing the old 'frontage vs depth' question. In the end the answer we found the most elegant and playable was simply to allow models to fight if diagonal coners touch; in effect this means that a wider unit will get an extra model or two into thefight, and more if it can fight in more than one rank.
The only place where the question of if a model can fight 'diagonally' has ever been addressed in is the Warhammer Fantasy rules, where there is a diagram that shows that models *can't* fight diagonally (or at least there was in the 4th edition). We did have a *different* diagram in WAB that showed that models *could* fight diagonally, but in the interests of consitancy this was dropped from the final book.
Whatever, I'd recommend players allow models to fight diagonally, as it makes the issue of width vs depth much more interesting, and actually is *literally* what the rules say you can do (you must touch to fight, and a model does youch on the diagonal). This said, I'll think you'll find that at the end of the day the size of a unit is determined more by the models you have available than any other issue.
Jervis
PS. Although no one has said it yet, the new 'falling back in good order' rules make bigger units more attractive, as they are both more likely to break an enemy, and not be broken themselves.
The on-going debate on this subject has encouraged me to go back to my history books and have a look at some of the formations that 'real-life' ancient units fought in.On the whole what evidence there is shows that 'heavy' units tended to form u into units that had roughly twice as many files as ranks. Occasionally the depth would be increased, so that the formation had as many ranks as files. Here are some examples:
The basic infantry unit of the Macedonian phalanx was the Syntagma, which was made up of 256 men that would normally fight in 32 files of 8 ranks, though this was quite often doubled up into a 16 by 16 formation.
Greek Cavalry fought in square shaped formation that was 16 files wide by 8 ranks deep.
Republican Romans fought in centuries of 60 or so men, and usually formed to fight 10 wide by 6 ranks deep. Imperial Roman centuries generally consisted of 80 men, formed 10 files wide by 8 ranks deep.
The points is that ancient *units* tended to fight in rather square shaped formations, while ancient *armies* tended to fight in long thing lines made up of these blocky units arrayed side by side. It's therefore rather dangerous to assume that forcing units to adopt a long thin formation is more 'realistic' in some way. The other point I'd make is that the weapons used by troops seems to have had very little effect on the number of ranks they fought in; this is because even the very longest pike could only be used over the heads of 4 or 5 ranks of troops in front, and therefore all of the really 'deep' ancient formations clearly worked on a psychological rather than a physical level. (BTW, the formation used by Greek cavalry does rather imply that I've been rather harsh in denying cavalry a rank bonus. I'm not actually bothered about this at all, as it stops cavalry dominating the game too much, but it's interesting to note that cavalry actually fought in rather 'square-shaped' formations too, rather than in thin lines).
All this said, it does seem to have been rare for a unit to *fight* in a formation that was deeper than it was wide. One way of addressing this would be to say that units that adopt a formation that has more ranks than files is treated as a march column. As march column's don't get a rank bonus, this would save such deep formations for battle-field maneouvering rather than fighting, as seems to have been the case, and would keep a lid on some of the more ahistorical formations that players use. A unit of twenty models could therefore deploy 5 files wide and 4 ranks deep if it wanted a rank bonus, but not vice versa.
Jervis