That level of displacement is horrific and should not be tolerated even for the most beneficial projects*. The fact that it's been done in Texas is emblematic of a broken system willing to
sacrifice its most vulnerable citizens in that way. One of the benefits of transit over highways is that the vast majority of transit projects can be done with minimal disruption by (re)using existing rights-of-way, re-allocating street space, and going around or under important sites. While some property acquisition is inevitable for larger projects, it's not a feasibility factor for most.**
In this specific case, I just can't see Medford Square as such an important node that it would justify the cost of rail rapid transit, much less doing so by taking a lot of properties. It's rather smaller than Arlington Center or Watertown or downtown Everett which could all justify rail. If you look at the
2018 bus route profiles, you see a small bump at Medford Square, but neither the massive spike you see at Watertown on the 70 nor the sustained density at Arlington Center or Everett. It's also not something like Roslindale where it would cut significant slow mileage on buses for riders. It will be the outer end of the 101, and midroute for the 96 which will be anchored by rapid transit on both ends. (The less-frequent 95 and 134 also have potential regional rail transfer opportunities at West Medford and Winchester.)
In some future where Medford Square gets substantially denser and somehow the 96 and 101 are insufficient, and BRT on Mystic Avenue and/or Route 16 are insufficient, and rail is somewhere necessary, branching the Green Line under Winthrop Street would be the same amount of tunneling as the Medford Branch would require. If the Orange Line could be run so frequent that branching is okay (and I would guess the closely-spaced stations and curves under downtown would make that difficult), going to Overlook Ridge via either the Saugus Branch (branching north of Wellington) or Broadway would be a better proposition.
*Maybe, maybe, you could justify it for a massive HSR project where geometry and scale limit less-intrusive options like existing ROWS and tunnels, and where displacement is over a much wider geographic area that can more easily permit relocations, and where those displaced represent a wider cross-section of the population. Even the massive CAHSR project seems to have significantly fewer displacements - the SF-Fresno section is claimed to displace
82 residential units.
**Again, HSR is arguably the exception - when the above-linked article was written in 2022, CAHSR had spent several years and $1.4 billion on land acquisition to piece together a brand-new ROW.