I have no idea how, but you accidentally moved a piece of US 422/Penn Avenue near Sinking Spring, PA and superimposed it over a piece of nearby Woodside Avenue, which also ended up with a gap in it. I've fixed these, but I'm not sure if anything else was accidentally messed up in this changeset.

Do not map solely based on satellite view. It's clear that you did because the US 422/Ramona Road intersection was reconfigured within the past few years. Most imagery available in the various OSM editors is too old to show this. When something is mapped with great detail, yet is radically different from satellite imagery, the imagery is most likely outdated.

I should also add that when a ramp splits again after diverging from the freeway mainline, the node of the subsequent split should not be given the same ref as the split from the mainline. It should be left blank unless specific exit numbers are signed at the split. (For example, a ramp signed as Exit 42A-B from the mainline, subsequently splitting into clearly signed Exit 42A and Exit 42B.)

Please don't add destination information to the motorway_junction nodes. Exit destinations are placed in appropriate tags on the ramp way (i.e. destination, destination:ref, destination:ref:to, etc; see osm.wiki/Key:destination). The exit_to tag was used in the past, but these days, the only things that typically go on motorway_junction nodes are exit number information and proper exit names. The latter are rare in America; the PA Turnpike's named exits on the ticket system are a prime example.

We do have an accepted official_name tag exactly for cases like this where something has an official name different from the accepted widely-used name.

That being said, with it certainly opening within the next week, I'm not sure it's worth reverting at this point considering how many small things need to change. I'll clean up your edit and mark it complete.

Source on the new road being open already? It definitely didn't open on New Year's and I find no news articles for its opening today.

I've reverted the southernmost portion of the US 301 spur road to highway=proposed when you had marked it as construction. The spur road was graded with construction of the mainline, but will be seeded with grass and not paved until the entire spur road is built, so it is not currently under construction.

I can't speak for the rest of the world (I know there are many trunk roads in the UK with no intersections), but motorway in the U.S. has been synonymous with fully controlled-access freeways since the beginning, as far as I can tell. Controlled-access means no intersections.

I-93 through Franconia Notch is absolutely dual-carriageway because the directions are separated by a physical barrier. It's no less dual-carriageway than the Pennsylvania Turnpike.

Tisdale is a freeway south of Gilcrease. Gilcrease is a freeway east of where it widens to dual-carriageway. The aforementioned segments should be motorway and the remaining segments of each road should be trunk.

Nope; it's just the first changeset, though I'm not sure how editing worked before the changeset system was added early on.

Hello! Thank you for mapping the last section of the Greensboro Urban Loop as being under construction. However, I'm curious as to why you completely deleted the existing highway=proposed ways, and remapped them with less detailed tagging?

Upon further investigation, you're right that the bulk of this segment isn't I-86 yet. Current signage at the I-81-NY 17 interchange to the east indicates that the I-86 section starts there heading east, but doesn't go any farther west. There is also no signage of I-86 immediately west of Binghamton, though new signage at the Prospect Mountain interchange does include I-86 everywhere, making for weird, spotty signage. The NY 7 interchange signage also features I-86, but only going west from it across the bridge.

Whoops, Source: TIGER 2016 wasn't meant to be added here.

I only did that in this one case because internally PennDOT keeps track of that portion of road as two separate one-way roads as if it were dual-carriageway, with one being SR 0011 and the other SR 8027. However, the road has no actual physical divider and is thus a single-carriageway road for purposes of mapping. If ref:penndot:forward/backward or ref:penndot:left/right are not best here, would "ref:penndot=SR 0011;SR 8027" be?

D'oh! Fixed...

I had some conflicts here because I accidentally clicked Cancel mid-upload. It finally uploaded seemingly alright with no validation errors in JOSM, but some stuff MIGHT be broken.

I was planning on having PA 8 marked as trunk up to Butler, because it's part of the NHS and has relatively high traffic counts, and even the occasional interchange. However, if this goes against common practice, I'll change it back.

Ah, the great "first."

Mid-upload my computer blue-screened and left the changeset open, partially uploaded. I tried to upload the recovered data layer (in JOSM) but got loads of conflicts, and attempts to fix them resulted in a LOT of duplicated ways/nodes. I think I got everything and the validator didn't complain, but let me know if I missed anything.