I don't know what the I2T (i-squared-time) response curve is for a normal breaker in an office, but I've used MCBs with trip delays equal to or longer than thermal breakers in MIL systems that had high inrush current and steady-state current greater than what thermal breakers can handle. CarlingSwitch was my favorite for COTS, Airpax for hard MIL.
There are a number of different types, I 'think' the standard one is type B? - when we moved the workshop at work a few years ago I ensured that the workshop was fed on it's feed from the main fusebox, with a wired fuse fitted (no way would I accept an MCB).
This feed goes to a fusebox in the workshop, with a number of different circuits, including an un-switched ring that feeds the computer, networking, and various infra-structure devices - plus a permanent power 'soak test' point. This circuit has it's own separate MCB, type B.
The rest of the power sockets are fed from a switched supply ring (using a cooker switch) and a type C MCB - even with type C it randomly trips a couple of times a year.
The lights are obviously on their own, are LED types, and fed from a type B MCB.
Basically there are light switches near the door, and the 'cooker' switch. When you enter you turn then both ON, and when you leave you turn them both OFF - thus making sure that everything (that you want OFF) is turned OFF. The computers etc. are unaffected, both by the switches, and by the tripping of any of the other MCB's.
I did later on add another un-switched double socket, for running anything else as well - mostly because I'd got a double socket going spare, and the type B MCB that I replaced with the type C one after the type B tripped every couple of days. It was trivial to fit the socket next to the fuse box, and run a wire from the spare MCB - a simple spur, and about a foot of wire.
Notice there's
NO RCD fitted, they have no place in a service environment.