Leave fortifying as it is.
Yes, fortification works fine.
Make building forts consume a worker but be built in 1 turn. Have the civ4/Thalassicus's idea of them being mini cities.
Being able to plop down a castle in the middle of a war, when the enemy is right at your door, seems absurd. Making them into impassible wars that will take several turns for the opponent to destroy, for the cost of a single worker, would be incredibly powerful. So in the middle of a war, I just bring workers along and all of a sudden my artillery unit that you're about to kill is massively protected.
Being able to hold the line on their own for a short while even without your 'standing army' supporting them; while also being a rallying when your main army gets there.
This is what cities are for. There is no way you should be able to do this for just the cost of a worker. If you want a long-term fortification that protects your unit and is hard to break into, then build a city.
The AI already struggles to capture cities; can you imagine how much worse it would be if you could add extra cities that it needed siege units to take (or melee units taking lots of damage) in the way?
The insta build by a worker means that they can be put up quickly in emergencies in war
Why is this desirable or realistic?
How the AI could be taught to use this is beyond me though.
Which is a dealbreaker.
* * *
If, for example, fort gave 2 gold base, like a village, but didn't get any of village's boosts
Then I would spam them in the early game to put my archers in them and make it even more impossible for the AI to attack me successfully, and then by midgame (before Economics), when I've already built improvements anywhere, I can use my worker downtime (saving them for railroads) to replace them with trading posts.
Same is true for roads. You might have a castle have the same build time as a road, not 3 times more than road (or whatisit).
This makes no sense to me. Castles usually took a decade to build. A small stretch of road didn't.
But if they're not useful, then they don't do this enough.
They are useful already, in rare, situational circumstances.
The problem with forts is not that they don't give a big enough boost, it is that wars don't usually take place in a small predictable area, and if they do, it is near your cities, and you want to be able to work those tiles.
AI fortifying cities with forts that have additional defenses against ranged attacks would incline the human to think more varied strategies than the outmaneuver +bombardments galore.
Fortifying cities and protecting units inside them is already modeled with the Walls/Castle/Arsenal/Military base building line.
If the AI had stronger defensive bonuses, then this would just encourage you to focus *even more* on lots of bombardments, because attacking with melee troops would be even more costly, why ranged attacks would remain costless.