Why the D Branch is light rail: It was built as inexpensively as possible, so as to add service quickly and cheaply. It was done at a total cost of $9.2 million, slightly under $70 million in today's dollars. The ROW was bought from the B&A for a song, rails were reused, stations were bare asphalt, the only big parking lot was built in a sandpit. A quarter-mile of cut-and-cover tunnel to connect to the Beacon Street tunnel was vastly cheaper than any possible route to downtown. Some aspects of that worked better than others. The level junction has always been a bottleneck, and the cheap stations and reused rails needed a major renovation within 15 years.
Subway conversion: the D Branch is unlikely to ever be converted to heavy rail. For one, you need Riverside as the main Green Line shops, with the ability to move equipment 24/7. There's no place on the system that's viable as a replacement. For two, it doesn't have the ridership to justify the expense of heavy rail conversion (including much larger stations with overpasses) - while it's a very productive light rail line, the Newton villages are never going to see the kind of development that would massively increase ridership. For three, the only likely heavy rail line that could get there is the Blue Line. If it's extended to Kenmore (perfectly reasonable), either heading west along the Pike or to Watertown and Waltham, or south to the LMA, would be much more valuable extensions.
GLX mode: An Orange Line spur along the Lowell Line to Woburn - with a possible spur from Somerville Junction to Alewife and Arlington - was in fact considered in 1962
. But the fundamental problem with that is branching - you're splitting northside frequencies between two branches, meaning that the major bus hubs at Sullivan, Wellington, and Malden only get half the Orange Line service. A two-car train of Type 10s every 5 minutes can move more people that an Orange Line train every 8 - and you can through-route a second Green Line branch to Medford if you need extra capacity. The Green Line also provides one-seat rides from Somerville/Medford and East Cambridge to the major job center of the LMA, which the Orange Line was not.
Extension of the Blue Line from Bowdoin was considered early in the GLX planning process. However, it would have cost about double that of extending the Green Line because of the need for substantial tunneling - either the Red-Blue Connector plus a new Charles River tunnel, or a very difficult (and maybe not even technically feasible) routing to reach the Lechmere Viaduct. Again, there wouldn't have been any capacity difference, stations would have to be much bigger, and you wouldn't have one-seat rides to the LMA or even Back Bay.