Will adding new questions to an interview discussion guide, halfway through a round of discovery interviews, cause potential for misaligned results at the end?
In most situations, yes. You'll need to measure this in some way, so adding more (or changing) questions will obviously provide different results.
However, this doesn't mean it's wrong to do this. Any procedure that helps you get better insights is a good procedure. If you feel like some questions need to be added, then go for it. Just be careful not to mix results from your first interviews and your second batch. This is specially needed if these new questions are going to have some effect over answers, so to save you time, just consider them as separate tests and combine results afterwards.
As long as you have a justified reason for the change- I would say it is okay to make changes. And once you have made the change, keep that consistent for the rest of the cycle of interviews. Have you done a pilot before you started? I found that a first couple of rounds of interviews (pilot) with the guide always pays dividends.