As it appears, 48.0 moves may indeed be optimal. Here are more 48.0 moves proofs, sourced from rec.games.chess. The PGNs were sourced elsewhere.
[Title "Andreas Rein, Die Welt 1989"]
[FEN ""]
[startply "96"]
1. h4 g5 2. h5 g4 3. Rh2 g3 4. h6 gxh2 5. g4 e5 6. g5 Ne7 7. g6 a5 8. g7 a4 9. g8=Q a3 10. Qg2 Bg7 11. f4 c5 12. f5 c4 13. f6 c3 14. fxe7 f5 15. b4 Kf7 16. hxg7 h5 17. b5 h4 18. b6 h3 19. Bb2 f4 20. e4 f3 21. d4 d5 22. Nd2 Bf5 23. Kf2 cxd2 24. Kg3 Kg6 25. c4 Kg5 26. exf5 e4 27. c5 axb2 28. Bc4 dxc4 29. a4 f2 30. Qdf3 Ra7 31. bxa7 b5 32. a5 b4 33. a6 b3 34. c6 c3 35. c7 c2 36. d5 e3 37. d6 e2 38. d7 hxg1=Q 39. gxh8=Q bxa1=Q 40. axb8=Q h2 41. f6 b2 42. f7 h1=Q 43. a7 b1=Q 44. e8=Q Qdf6 45. c8=Q c1=Q 46. d8=Q d1=Q 47. a8=Q e1=Q 48. f8=Q f1=Q+
[TItle "Theodor Burian, Rochade Europa 11/1997"]
[FEN ""]
[startply "96"]
1. b4 a5 2. b5 a4 3. b6 a3 4. Bb2 axb2 5. a4 Ra7 6. bxa7 b5 7. h4 c5 8. h5 g5 9. h6 Bg7 10. hxg7 h5 11. g4 h4 12. a5 Rh5 13. gxh5 b4 14. f4 e5 15. f5 d5 16. f6 Ne7 17. fxe7 f5 18. h6 f4 19. e4 Bf5 20. exf5 Kf7 21. f6 f3 22. h7 g4 23. a6 g3 24. h8=Q g2 25. Kf2 h3 26. Kg3 c4 27. d4 c3 28. Nd2 cxd2 29. c4 b3 30. c5 e4 31. Bc4 dxc4 32. d5 h2 33. d6 f2 34. Qdh5+ Ke6 35. a8=Q gxh1=Q 36. e8=Q+ Qe7 37. a7 e3 38. d7 hxg1=Q+ 39. Qg2 c3 40. a8=Q f1=Q 41. c6 e2 42. c7 bxa1=Q 43. cxb8=Q b2 44. d8=Q c2 45. f7 b1=Q 46. g8=Q c1=Q 47. Qgh7 d1=Q 48. f8=Q e1=Q+
In a comment on the answer to the PSE version of this question, linked in the other answer, user @Dennis Jaheruddin came up with a provable lower bound for this challenge.
"It takes 80 moves to move pawns forward. Obviously, they cannot capture the king and queen so there are only 6 promotion fields left on each side. Basically, this means it takes at least 4 half moves to clear additional promotion fields. Then there are 8 pawn rows that block each other, each conflict will need to be solved by capturing at least 1 piece that is not in its starting position, hence we need another 8 half moves. Now we already have a lower bound of 92 and there are still some issues to solve. As such 96 is probably optimal."
roving that more four half moves are necessary would be a momumentable achievement, finally bringing the 18 queens challenge to its knees. However, it seems no one knows where to start. Neither do I. As such, we are stuck at 92/96 proven plies. Hopefully, one day, a genius shall shine their light in our favor.