I think one of the biggest arguments against doing that is that when you read in our docs or watch our videos and they describe the keystrokes, then most of it won’t be true for you because you’ve mapped those keystrokes to something else. Lots of our keystrokes have a number of extra parameters, and it’s not trivial to set those up. So you’ll end up in a state where (ultimately) you won’t be able to use Dorico fluently because you’ll have incomplete or inconsistent keystrokes. Many of Dorico’s keystrokes don’t map onto a Sibelius equivalent.
I would advocate thinking of Dorico and Sibelius as very separate applications, and learn Dorico’s own keystrokes. It will be slower at first, but as you get more fluent, muscle memory will kick in and you’ll get the benefit in the medium term. Otherwise I think you may have the temptation to try to use Dorico like Sibelius, and that will be frustrating, because you won’t get the best out of it.