There are a few I'd agree and others I disagree with, for different reasons, and others I'll add.
Siege of Constantinople as in the final campaign within the Rise of the Ottoman Empire series? On Normal, I'd say that it wasn't necessarily tough, especially if you play Conquest enough to simulate fighting on different fronts to achieve an objective, and get used to Petain's strategy of swapping exhausted legions out for fresh ones.
Battle of Demetrias was tough because the Venetians had a cheat on their end, but I think I resolved it with tech and troop upgrades, as well as a strategy based around the somewhat obvious trick involving the reinforced gen.
Arguable, but worth noting.
Harrying of the North post-update (which removed the OP Templars) was tough, especially due to the horrible terrain, trick gimmick involving the Danes, and the difference between friend and enemy AI.
Notable difficulty, doubly so because of my style.
At the time, I'd agree.
Battle of Demotika was really that bad. I don't think anyone would disagree, even if you had IAPs.
Insufficient technology per turn, being stranded across the Dardenelles, and having to rescue a city poorly defended while the enemy has all possible advantages, are just the right conditions for this.
The gimmick wasn't worth the effort in the end, right? Reinforcements still showed and the morale drop wasn't too significant. The Palaiologos forces still attacked.
Battle of Pelagonia:
- I think that once we found out the trick, it wasn't so hard, whichever strategy we chose (trap the gen or chase him away).
The clearest difference are the Epic Battles.
I've yet to clear Hastings on the Norman side due to a lack of higher-level troops and silver coins for the gold equipment (they are dispersed for Territory and Expedition), Tang at Talas (due to lack of clarity on the betrayal and troop levels), and Crusaders at Hattin (time and troop levels).