I'm going to say, the two posts above this one, if taken together, scare me considerably.
The threshold armor method has a critical failure, in that when weapon damage and body point values go skyrocketing, armor becomes a non-issue.
However, the %reduction method has a flaw as well: it makes all weapons cale the same, so that (to coin an MMO's phrase), the DPS of a weapon is it's single most important factor.
The machine guns, as was stated by the post 2 up from this, have one of the better damage/time ratios in the game, when not counting armor. If armor is a straight %-reduction method, then whatever rankings the weapons fall into for sustained DPS will be, directly, their exact usefulness rating, witht he only variable being "body type" bonuses; if I remember a previous dicussion on them, however, I was told that their % bonus is applied BEFORE armor calculations, which (when dealing with suitably large %'s, like 78% for dragon reduction), makes their effect far less useful towards the end-game scope, which is where the original threshold's method falled apart as well.
Machineguns, no matter how advanced, should not be decimating Dragon bodies. 1 damage per hit is IMO the ideal for that type of confrontation; IIRC, a fully upgraded Assault gun (as you've mentioned the T-A-G is now just 2 A-G's firing in tandem), does 68 damage. -78% for dragon armor, and that's 15 dmg per shot; 2 shots at once, and a dragon body takes 30 damage per hit. Given the ROF of the lategame machineguns, a T-A-G would win a 1v1 encounter with... heck, just about anything.
I must admit that the current setup does make a kind of sense - if you think of a modern day tank you really aren't going to do any damage to that with a machine gun.
A gatling-cannon, like that on the A-10, however, is specifically designed for taking down tanks. So there should be some way for low-damage-per-shot weapons to have an appreciable effect on armoured targets, which is why I suggested the option to have a percentage of penetrating damage that ignores the armour calculations.
I'd like to play WZ with the %-based reduction before I make any further judgements - intuitively I expect it to reduce the variability across weapons - upgrades would be less essential and older technology would fair a bit better against an quicker researchign enemy. Whether that is what will happen I'd have to play to confirm, however.