Even though MD 53 is not signed here on I-68/US 40, I think the 39 label should be synced with MD 53 since it has a ramp to I-68/US 40 EB.
Agreed.
Also I thought that MD 658 would have a separate point for MD 49 based on what I see here with PA 42 at I-80 and PA 44.
Hmm. Normal policy is "one point per interchange" - I would think that when two routes intersect within (or at least at the geometric edge of) an interchange it therefore logically follows there should only be one point where everything meets up. After all, doing it the other way would mean having a segment of MD 658 between I-68 and MD 49 which creates some statistical absurdity. Someone who came south on MD 658 and made a left on MD 49 would claim a different length of the route clinched than someone who came south on MD 658 and made a right onto I-68 WB, despite those two turns being at the same intersection. Meanwhile someone coming from MD 49 can proceed straight onto I-68 WB (and someone getting off of I-68 WB can proceed straight onto MD 49) without traveling on any of MD 658... so it does make sense for MD 49 and I-68 to be shown as intersecting.
On the other hand - yes, MD 49 and MD 658 intersect at a different point than MD 658 and I-68 cross, even if the former is within/at the edge of the interchange footprint of the latter. Mapping it this way would better approximate the actual physical anatomy of the roads involved.
I am inclined to favor the first method, and perhaps suggest PA 42/44 should be rejiggered to also follow it. But I'm open to being convinced otherwise.
Also note to self for later - check for further precedents in other places.
