I know you believe, John, that Dorico is conservative at heart, and that by design it discourages innovation. I would disagree with that characterisation. I feel as if I am repeating myself, but since you are too, I guess that’s OK. Dorico is intended to be very flexible and customisable (it provides literally thousands of options for customising the behaviour of the application and the appearance of the notation it produces).
But it is also intended to be semantic where possible, i.e. to have an understanding of the intention behind the notations that are being used, so that it can help you when you need to perform tasks that would otherwise cause you to have to do a great deal of manual work. Some simple examples of this: the fact that music belongs to instruments rather than staves, so you can trivially move all of the music intended for one instrument from one player to another, either creating or removing a doubling in the process; the fact that all notes are represented by default in abstract durations so that Dorico is free to notate and re-notate them according to the meter and following your preferred rules for note durations; the fact that it uses rational numbers (i.e. fractions) to represent pitch, allowing you to set up complex tonality systems with arbitrary levels of precision for the amount by which accidentals raise and lower notes; the fact that every notation is handled semantically rather than graphically, so you can quickly and easily do things like increase or decrease dynamic intensity, or change the appearance of a hairpin to a textual form, or have Dorico help you identify notes that don’t fit with the prevailing harp pedals; and so on, and so on.
The price of this semantic approach, however, is that Dorico is only smart in the ways in which we have so far imbued it with smarts. Once you stray outside the areas we have focused on, you are more in the realm of workarounds that lack elegance and intelligence.
I would characterise the difference between Finale and Dorico that Finale has weaker musical semantics than Dorico. I think you can make the argument that the evolution of music notation software over the past half century has been towards stronger musical semantics, rather than going in the other direction. SCORE is essentially a graphics program specialised for music layout; the applications of the 1980s and 1990s have added layers of semantics (instrument ranges, checks that bars add up to the right number of beats, playback that approximates not only instrument sounds but also specific playing techniques for those instruments, etc. etc.); and Dorico, an application of the 2010s, further extends that approach to a much broader swath of what it does.
I would argue that an application like Finale is freer only in so much that it understands less of what you are trying to do. Things perhaps don’t feel so much like workarounds in Finale because everything on some level requires manual work in Finale.
We absolutely want to put more and more power into the hands of our users as Dorico matures. But we believe that power should, if possible, be smarter than simply allowing you to do all the manual work yourself.