On reflection, I think the 1e WvAC is a bit fiddly and a poor simulation, which makes it not worth it. The fixes I can think of to make it a good simulation are quite fiddly, and probably not worth it. So I end up preferring the solution which is not at all fiddly and doesn't try to simulate it at all.
The shield is the biggest problem. Shields are already way undervalued in every edition, so whether you have a shield or not should make a big difference. But then you need to simulate both armor type without shield. Maybe you could have a separate column for shields and sum the armor type adjustment and the shield adjustment, but the existing AC 9 column is the opposite of what I would want to see for that.
It would help to shorten the categories to hacking, slashing, crushing and piercing weapons, but you would still have the shield + armor issue. In that case it might be simpler to place it in the monster stat block (different AC for hack/slash/crush/pierce) then to have players track the adjustments for shields plus 9 different armor types (since padded would be different from leather, and studded leather might be different from ring).
How can you not love the notion of a druid tap-tap-tapping away at a fighter in plate mail with his quarter-staff? (or on a dragon's hide?)
See, this is part of my issue.
Staves are greatly underestimated, and two-hand swords are greatly overestimated. A quarterstaff is not a broom handle, it is long, heavy, fast, accurate, and wielded in a manner that takes full advantage of using the muscle in both your arms and the weight of your body behind them. And unlike steel it doesn't bend. It won't cut through plate, but neither will a sword, it is all blunt force trauma.
A two-handed sword, on the other hand, is lighter than you think, and while it employs leverage in the same way as a quarterstaff it is quite flexible and transfers energy less efficiently. The edge won't cut plate but will cut flesh very nicely. Having a staff be
better than a two-handed sword against an unarmoured opponent but
7 points worse against plate is ass-backwards to a phenomenal degree. Also, they should probably have similar speed factors.
It is particularly stupid for a staff to have a penalty against chain. The rigidity of plate will spread the force of a blow over an area, but the flexibility of chain doesn't do it at all. The compression of links might soften a blow somewhat, but I doubt it would be enough to be noticed in a system that measures AC in 5% increments. Basically, chain (which is assumed to be worn over padding) ought to be the same as padded armor; if we accept that a staff is +1 against padded armor (which is also debatable, but leave it be for now), then it should be +4 (or +3 if we are generous in our assessment of ring compression) against chain, not -1.