I personally don’t think it’s an extra step
But it is a different order you have to be aware of.
Each with advantages and disadvantages depending on what you do.
You can sort of see this as one way will set in-point in pre-twixtor time and the other in post-twixtor time. Frankly, we would prefer if both these apps allowed us to extend the duration of a clip within the plugin then we could make this more efficient. Actually from something like CS4.0.2 to CS4.1 Premiere did it the way you describe FCP does and many users complained…
You can sort of see the Premiere workflow as requiring you to prep a sub-clip if you like (i.e. isolating the section you want to slow-mo). You can then for example set Twixtor on the clip on a video track and set in and out point to insert in your timeline so it fits a particular duration, and change your mind, grab the in-point and change it and the other frames don’t move…
In FCP if you want to change the in-point after applying the effect everything moves which does not work well if you are in Twixtor Frame Mode, e.g. need to synch to an audio event… Your preferred way might be more convenient in the case where all you do is set a constant speed value in Twixtor as you then set the start point directly when you set in point UI.
Pierre