This discussion of the value of tank in a combined arms operation in an era of proliferating weapons that can destroy tanks notes this issue with fighting in cities:
Fighting in Mosul has shown that an insurgent force can maul and disrupt an organised opponent if given the space and time to prepare.
In this 1997 Land Warfare Paper about Iraq's 1980 invasion of Iran I wrote:
The demonstration that troops apparently hopelessly outclassed can make a good showing - even if they have to do nothing more complicated than die in place in their bunkers - is useful. Iran's ill-coordinated light infantry forces were stubborn obstacles to Iraq's ambitions when deployed in the cities of Khuzestan. Fighting a determined foe block by block and house by house as the Iraqis did in Khorramshahr would force our Army to play by our enemy's rules. Although it is possible that information dominance could extend our superiority in open warfare to urban areas, that breakthrough has not happened. We must not forget that urban conditions may limit our technological and training advantages, lest we experience our own Khorramshahr debacle one day. [emphasis added]
The battle for Mosul against fanatical jihadis took a tremendous amount of time and inflicted serious casualties on the best Iraqi troops sent in to dig out the ISIL forces despite massive amounts of Coalition surveillance and precision strike support. Urban conditions absolutely limited technological and training advantages nearly forty years after that Khorramshahr lesson.
And
that leveling aspect of cities is one reason why I wanted the Army to
try to "bounce" Baghdad on the run during Iraqi Freedom before Saddam
could organize and mentally prepare for a defense of the city. In my
post on Iraq on the very first day of my blog's existence, July 12,
2002, I wrote:
The only potential battlefield problems are city combat in Baghdad and chemical weapons. With his regime at stake, die hards may well pull a Berlin bunker strategy and Saddam may figure he has nothing to lose by going chemical. For the problem of city fighting, remember the 82nd Airborne should be on call to aid in this. Hopefully some defectors will aid too and lead their own Iraqi troops into the city. In any case, we will need to drive on and bounce the city rather than besiege it and give the defenders time to fortify their nerves and buildings. Better some losses early than a dragged out bloody fight in that sprawling city with the accompanying CNN risk of starvation and disease plaguing the civilians.
The Army and Marines in 2003 did what I hoped--hit Baghdad while the enemy was reeling. I had thought we would rapidly advance to the outskirts of the capital although I did not predict the bounce would work. I just said it was worth a try given the alternative of giving an enemy the time and space to prepare.
Cities can be taken without heavy losses, as America did in 2003 (although holding it through the Surge and Awakening successes was another matter, although successful, too). And sometimes you must take a city as Iraq had to liberate Mosul from ISIL in Iraq War 2.0. But in between you can have high-casualty operations to take cities you don't really need to occupy.
I remain concerned that the Army may be too eager to plunge into megacities. Capabilities to fight in cities don't mean we should fight in cities.