That's HTC nr.25, and again (as in nr.26 in the other thread) it seems to be an idea to analyse, to delay the given solution for one move by 1.Rad1.
Again Cfish with Pure Eval of the same latest net from here
Depth 41, again after Forward- Backward of both lines:
1rb1r1k1/2qn1pbp/pp1p2p1/2nP4/P1p1PP2/2N1BQNP/1PB3P1/R4R1K w - - 0 1
Analysis by CF EXT 270820 x64 POP N:
1. +- (4.04): 1.e5 Qd8 2.Rad1 dxe5 3.f5 e4 4.Ncxe4 Nxe4 5.Bxe4 Ne5 6.Qf4 Bb7 7.Bd4 Qd6 8.Bf3 Rbd8 9.Ne4 Qb8 10.fxg6 hxg6 11.Nf6+ Bxf6 12.Qxf6 Qd6 13.Qh4 Rd7 14.Rfe1 f6 15.Qg3 Rde7 16.Bxe5 fxe5 17.Rd4 Rf8 18.Rxc4 Rf6 19.Rg4 Bc8 20.Rh4 Bf5 21.Qg5 Rff7 22.Be4 Bxe4 23.Rexe4 Rf1+ 24.Kh2 Rd1 25.Rh6 Rg7
2. +- (3.30): 1.Rad1 Nd3 2.Bxd3 cxd3 3.Rxd3 Nc5 4.Rdd1 f5 5.Bd4 Nxe4 6.Ncxe4 fxe4 7.Nxe4 Bf5 8.Ng3 Bxd4 9.Rxd4 Rf8 10.Re1 Rbc8 11.Kh2 b5 12.axb5 axb5 13.Nxf5 Rxf5 14.Qa3 Qc5 15.Qd3 Rf7 16.Re6 Qc1 17.Rxd6 Rc2 18.Rc6 Rxc6 19.dxc6 Qxc6 20.Kg3 Re7 21.Rb4 Rb7 22.Qd8+ Kg7 23.Qd4+ Kh6 24.Qe5 Qc2 25.Qe3 Kg7 26.Qd4+ Kh6 27.Qe5 Qf5 28.Qxf5 gxf5 29.Rd4 Kg6 30.b4
So this time the two moves might be to near to each other to give a single one solution only points for finding maybe, even if 1.e5 has the edge yet.