That's a huge difference in case of a fire. Ninety seconds of smoke filled, oxygen deprived, or hot air, and many people will make it out alive. Three minutes and most will have passed out.
There's quite a few instances of burning planes being evacuated where not everybody, but a sizeable portion, made it out. Seconds matter*.
* Almost always there was some idiots slowing everything down by grabbing their luggage.
Here's a good example because it has a clear cutoff point: "Less than 90 seconds after touchdown, the interior of the plane flashed over and ignited, killing the remaining 23 passengers on board, who died from smoke inhalation and burns from the flash fire."
https://en.wikipedia.org/wiki/Air_Canada_Flight_797
It seems obvious that a faster evacuation would have saved more people and a slower one less, roughly proportional to speed of evacuation.
But - the test handles this by only giving them 90 seconds. In the real world it would take far longer, but rather than testing a longer, more realistic test, they do a really fast one instead.
If the happy-path evacuation isn't possible to begin with, e.g. for design reasons, the more complicated realistic ones will be negatively affected as well.
I would expect this to eventually be rolled out more widely and later replacing real windows, which are structural weak points of the hull and heavy. Also, for our Ryanair no-frills flyers: they are screens where you can show ads.
For the Emirates experience, see https://www.youtube.com/watch?v=VdDzaNlGSmk
Also, anecdotally from my flying (I'm a private pilot, small, single engine planes) turbulence tends to create much more vertical translation and pitching movement than it does roll movement.