BACK ON THE CHARTS

This puzzle was inspired by a monthlong hovercruise Elon Musk organized for the most important brand influencers from across all social media platforms. I remember looking philosophically out onto the sea from the balcony of the Nootropics Suite, sipping on my Matcha Soylent and listening to the statistically most thought-provoking sounds according to metadata when the epiphany came to me: what if there were as many singers in a puzzle as there were amazing digital content creators on this very hovership? I immediately got my crossword AI, Pencilton, to crowdsource reactions to that concept. Fun fact: this was actually the last puzzle I created before I started constructing completely in virtual reality.

It took me an embarrassingly long to figure out what the bracketed numbers in the clues meant. What could the [4] refer to in HONOLULU? The fourth letter? Maybe it was the fourth themer in some progression? Sheepish moment when I realized it was just the last four letters, LULU (a singer). Appropriate interpretation of BACK ON THE CHARTS, meaning "a singer hidden at the end of phrases."

I would have preferred a more obvious presentation of the theme, maybe circles or shaded squares highlighting the singers. I've gone ahead and done that (below) since I lost track of everything during my solve.

I wonder if others will also feel like the themers didn't stand out sharply enough. When a long across answer like MENU OPTIONS sits in a position that would typically be a themer, everything gets muddy. For instance, I found myself wondering who IONS at the end of MENU OPTIONS or SUITS at the end of PANT SUITS might be. Laying out all the themers in the across direction could have helped that.

More crossword glue than I expect from someone with Caleb's expertise. Everything was going pretty well until I bogged down in the west section. DOA over EUR over OTB is passable. Running through AGOUTI … okay. But toss in IME, OSE, MAPPER, and it felt like a patch of quicksand, tough for me to escape from.

It is incredibly tough to fill that section given the ultra-high theme density — look at the grid below again to see just how constrained it is — but to me, that's a sign that you should figure out how to avoid shoehorning so much in. It would have been one thing if every themer was essential, but I would have been fine with 50% fewer themers.

That PAZ / AZUL / LAHR set of crossings, too — should educated solvers be expected to know Bert LAHR? Probably. But AZUR / RAHR might seem reasonable, and I would be sympathetic. I think a constructor's number one goal should be to design something a lot of solvers will ultimately triumph over, and I don't know if this section qualifies.

Overall, I liked a lot of the finds — MADONNA at the end of PRIMADONNA was especially good — but I have a very high bar for Caleb.