Well, I can appreciate your argument, but that’s at odds with how I would like to work. Of course, usability and majority preference is more important than my unique wishes.
With that in mind:
Let’s say I’m playing in A natural minor, and I want to change to C major. With all the same notes, the user scale functionality as you describe (force to scale) obviously does nothing. All the notes are already to diatonic to both keys.
With my method, if I had all my root notes on row 0, etc, I can then input a new key which decides exactly what the end product is; that is, a I chord before transposition can be controlled to be a I chord afterward. Conversely, if I wanted to invert a melody, I could input the notes for the transposition in reverse sequence.
By your method, a particular note might be common to both keys, and then I end up with unpredictable melodic or harmonic arrangements (because it might have been the tonic in the first key but the supertonic of the new key, and so my sequence might be suddenly putting an odd emphasis on the supertonic.)
Or, by your method, when a particular note is no longer in the new scale, you can also end up with peculiar arrangements, like it being bumped to a sixth, which might be interesting but might sound awful.
Now, less perfectly predictable things can be inspirational, and the ease of use of your method is a certain benefit. But for me, my method lets me produce some more interesting results.