Thinking big picture, changing to a more agile technology for this corridor is a money saver. Outer sections can be constructed for less, and that can offset the cost of refitting the existing 5.5 km.
But if one looks at the next step alone, then the long-term saving from the more appropriate technology is a hypothetical future benefit, while the cost of conversion is immediate. Let's assume they just want to extend from Don Mills to McCowan as the next phase. That project alone is cheaper if they retain the existing technology, and that's what they will likely choose.
But then we will have a 14 km-long line that uses the existing technology. Now, any conversion is both more expensive (more km) and more tedious (temporarily losing a line that has gained a greater usage) than if that was done while the line was just 5.5 km long. Thus, any new extension uses the existing technology, or the extension doesn't happen at all.
That's what I mean talking about the government chosing the path of least resistance (and thus delivering suboptimal results in the long run).