I managed my time very poorly in the AM and got, optimistically, 24-28 I'd say, including educated guesses (as opposed to WAGs) as perhaps 50% chance of being correct. The last 8 questions left I answered pretty randomly in the AM, and many before that were rushed too. Oh yeah, also my Casio (which I hadn't used since taking the FE exam 10 years ago: for work I use an illegal TI-86 or, you know, MATLAB) somehow got stuck on fractions-only mode and I couldn't figure out on the fly how to get it to output decimals. I ended up multiplying all my answers by 1.000001 just to get a real number instead of 342/7 or whatever.
The afternoon felt much better, both because I had time to consult my manual about the calculator over lunch, and because I got off to a much quicker start for whatever reason: although I also ended up answering the last 8 or so questions randomly, I felt I also did better on the questions I was able to get to. Maybe 28-32 correct in the end on the PM. So, summing up, I'm anywhere from a definitely-failing 52 to a definitely-passing 60. A nailbiter.
Biggest area of weakness: anything to do with a thermodynamic cycle. I only used the steam tables for like 2 problems. It's a good thing all those power cycle problems were at the end of the PM section because I didn't really know how to do them and I ran out of time and ended up guessing on most.
(P.S. I didn't study at all--literally zero; I read the intro to the MERM the night before the test to figure out what was covered in the book--and only brought in the MERM as a reference. This was probably not the best course of preparation. If I fail and decide to retake in October, before the CBT switch, I'll actually do some studying.)