Taking a lightly defended city isn't a big deal - one stack is as good as another for easy targets. I'd make it a city with 60% cultural defense, walled, on a hill, that can produce another longbow every 2-3 turns. The real advantage of cho-ku-nus is that you get them before Maces and Trebuchets. A "more realistic" test would have the maces and trebs facing a slightly more advanced and better garrisoned city. A decent sized stack for the era would have around 10 attackers minimum, with 4 accuracy promoted catapaults. Ideally they are not attacking at all - they're there to remove city defenses. Trebuchets get city raider promotions, not barrage. The point of the test (I assume) is to take a highly defended, strategically important city and be able to continue the campaign. It's not to simply take out one city. You're likely to face 5 or 6 longbows and maybe a few axes/swords or regular archers. If they have catapaults of their own, the drill-promoted Cho-ku-nus are at at further advantage since they are resistant to collateral damage. Entering borders at a 60% defense city means crossing at least two tiles, maybe 3. That means expect at least one more whipped unit and maybe reinforcements coming via the roads. So make it 6-7 longbows, with another coming in 2 turns if the city isn't taken. I don't think the test would have to be done 20 times - 3 or 4 times would be enough to give a good indication of which stack preserves more hammers.