I'd also say Exit 69 is within the confines of Exit 68, and thus, single interchange.
See also:
CT I-291 1.
OK, and I-95 has one interchange (79) there, as does I-195 (unnumbered according to GMSV) <shrug>. Regardless, the intersection is badly misaligned, with a segment of I-195 left unattached to anything at its north end, which is what initially attracted my attention.
I would say, go with two points on this one.
WRT graphing algorithms: VA I-64 186 & VA I-195 I-64 are at nearby coords; assuming it's close enough to count as a junction, routing stuff between the three roads could still work here. Though from a visual/graphing perspective, the hanging end of I-195 is a bit unsightly.
Seeing a similar hanging end on NY I-278 is what prompted me to start a cleanup operation there.This reminds me of the
I-96/696/275/MI5 junction mess I tackled a couple months ago. This looks like another case where it's darn near impossible to do right.
If I were drafting this all now... I'd leave the 64/195 junction where it is. VA I-95 79 (and thus VA I-64 187), I'd put midway between the two sets of flyovers, on the I-95 centerline. (This does break with Tim's "middle of the central ramp triangle" guideline, but I don't always believe that's the most appropriate thing to do in all cases.)
As for the end of I-195? I dunno man, that 90° turn toward I-95 might look a bit ugly. I could do that, or I might be inclined to do something a little similar to I-275 from the Michigan example above, and end I-195 where it hits the interchange complex, at I-64, and consider the bits north of that as just ...ramps. It still works from a graphing perspective.
Wibbly-wobbly, rampy-wampy, STUFF...
This post is just for philosophical purposes, and isn't intended to advocate for or against editing or not editing any files any particular way.
BTW, I haven't actually looked at any of the stuff discussed in the first two posts, so...
PHEW!