I am still working on both transparencies (Kodachrome, shot many years ago) and digital capture via D200.
Yes, they are very different, to my eyes at least, but whilst I still have a film camera and a freezer drawer pregnant with materials, the simple hassle of spotting my scanned pics is just too daunting nowadays. The other huge problem is processing, where not all labs are created equal, and even the good ones can vary in output. Further, whilst colour transparencies and the E6 system are all well-known and can generally be monitored well enough by the labs, processing b/w film seems to be far more doubtful - possibly because the exposure of that material is so much more a matter of personal interpretation and one film with one developer, learned at home or in your own business, becomes a technique which no outside lab can handle in the same way.
I have also learned one of the newer bad habits that comes with the digital age: impatience. I have also lost a lot of my ability to use an incident light meter as well as I used to when no alternative was available; the matrix system in the D200 is just so good that I needle match all the time and never bother looking at the rear screen until I get home, and then only in order to wipe out the trash.
I looked at the Mamiya 7 a few times, thinking the Leica thoughts, only larger, but the same argument that kept an M3, 6 or 7 and myself apart still applied: non-reflex focussing is a step too far, possibly because I started off with a modest but to me, at the time, expensive Voigtlander Vito B which eventually carried a rangefinder in the shoe... The Exakta that replaced it made slr viewing irreplaceable, particularly as I became very fond of the 135mm focal length for much of my work. The Nikons that replaced that machine were so good - I used to imagine that had Nikon made a version of 6x7 then it would have been mine - sadly, it turned out to be a Pentax 67 that caught me, but that´s another tale.
Starting from scratch, it would have to be digital capture now.
Rob C