Underlying the "underestimated" line is the "estimation" valuation of Protective.
Lots of people value Protective as the least powerful trait or close to it. Of course, that depends on what it is exactly that you're trying to do. For some posters, opinion is that Protective is actually a handicap - which just isn't true.
Depending.
On its own, Protective certainly doesn't impress. CG1 and Drill1 combined make for a nasty defensive unit, but it doesn't attack anything. Players in SP games like to make their games on attacking enemies, so you can understand how many players would feel that attacking units are more valuable than defending units.
At the crux of the decision is the question: is it better to attack or to defend?
I would say that unless you know how to use passive defense as a leverage for a good game-winning leverage, then you have no business commenting on how good or bad Protective is.
What does that mean? Simply put, it means that there are cases where Protective is better, and there are cases where Aggressive or Industrious is better. Trying to use Aggressive to build monuments is colossally stupid - it's self-explanatory. Likewise, trying to use Drill3 CG1 Pinch Muskets to attack is also pretty weird.
IMX, there are some cases where you can't attack or counter-attack an enemy stack to any useful end whatever you do and whatever other trait you may have. This is especially the case when you're at a disadvantage going up against UUs, Civ traits, or UB advantages, as well as terrain.
It goes without saying that a Drill3 CG1 Pinch Musket is a god on passive defense. The question most higher level players would level is, "How can you use it to win?" Speaking only from my experience, I've use Pro Archers and Pro Gunpowder units together with the cheap Walls (and Castles when possible) in order to present a strong defense on cultural borders when necessary. Strong City defense of this nature is cheaper on hammers especially against the AI, and especially when you can position your defensive units to force the AI onto bad terrain.
Theoretically, you can eschew the defense altogether and instead simply create two offensive stacks and take over the enemy Civ using, say, Industrious Wonder advantage instead of Protective. That's not actually true. Walls with Stone (and even without) on Protective are so cheap that you can effectively spam it in every City if you need to. With even 2 Protective Archers (which you can shift around as necessary), that's such a staunch defense that you can basically allow it to be attacked by a much larger enemy stack and still come out ahead.
Too, moderately promoted All Drill defenses on a city is extremely resistant to siege weaponry and thus to attack stacks. While you can theoretically eliminate the attack stack or recapture the city with a stack of your own, that kind of hammer investment can be used for bigger and better things - like taking enemy cities.
I think Tokugawa's early game demonstrates best how the Protective trait can be used for hammer leveraging. Workers + roads + Swordsmen or Axes to invade neighbors, Protective Archers and Walls to protect early gains from enemy Civs and Barbarians without taking the focus away from your attack stack.
In theory, Protective can be used this way with any civ, but in practice, Protective can only leverage hammer gains when protecting assets under threat. Without an attacking trait or UU to provide a spearhead, it loses much of its luster. Sitting Bull's Protective Trait, for instance, is not as useful as Tokugawa's is.
The winning point for me was the observation that Protective is a monster trait in MP. If you can use it to incredible effect against a smarter opponent than a computer, then certainly it's going to be much more useful against a comp AI? Of course, one can argue that you can bamboozle the AI much more effectively using alternate traits, at which point the problem really isn't the trait itself, but the fact that the AI is a weak opponent in the first place.
