This grid is the answer to the question, "What if a themeless grid had not one seed entry, but four?"
This grid shape came about because I was determined to stack 3 of the ~7 killer 13s I had lying around. After that proved too hard, it morphed into two 13s and two 12s, separated by a more segmented middle row. But that was still too hard, so they just kept drifting further apart. Happy to report that I got to use my #1 and #2 favorite 13s (19A & 41A, respectively) and my #2 and #4 favorite 12s (38A & 21A). (More info underneath the puzzle!)
Gotta give shoutouts to Brooke and Paolo for test solving. I'm continually impressed by how thoughtful they are, both in the sense of having lots of in-depth and valuable feedback, and in the sense of being generous with their time, and very kind people overall! I think this puzzle might be harder than my first themeless, but it's hard to get a feel for these things when one of your test solvers finishes in 2:58. I'll let you guess who it was...
Finally, if 6D is unfamiliar to you, I'm gonna need you to watch this. (SPOILERS, obviously).
-SPOILERS-
There probably won't be many spoilers in this, but it seemed like too much to frontload at the top of the post, so I'm writing it here.
I still don't fully understand making a themeless puzzle with only one seed entry. There, I said it. Let me explain.
In my first themeless puzzle, I started with three interlocking 15s. The original plan for my second themeless grid was to try my hand at stack-building, which is kinda how this started, as alluded to above. I started with my favorite 13, LAYERS OF IRONY, and searched through my list of 13s for options to stack alongside it. It didn't take long for me to get really attached to a handful of about 5 entries I really loved, and I wasn't even interested in looking at other options. Before long it became apparent that I was no longer creating a triple-stack with one seed entry, I was creating a triple-stack with three seed entries. Which, of course, did not end in success.
From there, I tried introducing some 12s to the mix, figuring that if I couldn't have 13/13/13, maybe 12/13/[segmented center row]/13/12 could be good too. When this, too, proved too ambitious, the 12/13 and 13/12 stacks just migrated closer to the edges. After many days of cycling through combinations of words, pairings, left/right justifications, spacings, and black square placements, I finally landed on the arrangement you see in the final puzzle.
Crucially though, except for the very beginning, I was never "picking a seed entry and building around it," the way that you're "supposed" to. I was instead "picking four seeds from a list of 12 and finding a way to force them in." Honestly though, I think that approach was better than what I could've done by just building off a single entry. To put it bluntly, my construction skills aren't nearly extensive enough yet to end up with four long entries that I'm this proud of, unless that's my deliberate starting point. I still really want to learn how to build a themeless the "proper way," but for now, I'm very happy with the result I got in this puzzle.