"Portfolios for long-term investors" is published, Review of Finance 26(1), 1-42. (2022). This standard link works if you have institutional or individual access. I am not allowed to post a free access link here, but I am allowed to post one on my webpage where you will find it.
The theme: How do we account for the vast gulf between portfolio practice and portfolio theory? How do we make portfolio theory useful given that the world has time-varying expected returns and time-varying returns and correlations? I argue for a view more focuses on prices and payouts, as a long-term bond investor should buy an indexed perpetuity and ignore one-period returns. I advise one to think about the market and equilibrium. The average investor must hold the market portfolio. Anything else is a zero sum game. So figure out why you are different than average. (If you think you're smarter than average, note that they think they're smarter than you.) The result encapsulates some ancient advice: Buy stocks for the dividends, broadly interpreted. Take risk you are well-positioned to take. If stocks have great value to other investors for reasons either technical (liquidity, short-sales constraints, etc.) or behavioral, avoid them.
The paper grows out of the summary I used to give for MBA and PhD students. There aren't any equations, but there are lots of suggestions about how academic portfolio theory might be done better, and connect better to its intended audience. Thanks especially to Monika Piazzesi and Luis Viceira, who invited me to give the talk on which it is based, and Alex Edmans who shepherded it to publication.
Update: I just found that the NBER has posted the original lecture video