HI,
I checked the guiding pulse in the INDI logs: the DEC axis has 2073600 steps, which is roughly 24 steps per second for sidereal rate, and the 1.250 North guiding pulse at 0.5x shows a difference of 13 steps which seems correct. From the driver point of view, it can't do more.
I noticed in your first PHD2 logs some problems while calibrating. In the RA case, East/West moves seems ok (8 west moves, 4 east moves:

01:04:15.864 00.000 3013644336 MoveAxis(W, 1250, -)
01:04:21.199 00.000 3013644336 MoveAxis(W, 1250, -)
01:04:26.560 00.000 3013644336 MoveAxis(W, 1250, -)
01:04:31.903 00.000 3013644336 MoveAxis(W, 1250, -)
01:04:37.187 00.000 3013644336 MoveAxis(W, 1250, -)
01:04:42.493 00.000 3013644336 MoveAxis(W, 1250, -)
01:04:47.788 00.000 3013644336 MoveAxis(W, 1250, -)
01:04:53.049 00.000 3013644336 MoveAxis(W, 1250, -)
01:04:58.320 00.001 3015503888 WEST calibration completes with steps=8 angle=-178.7 rate=2.618 parity=1
01:04:58.344 00.000 3013644336 MoveAxis(E, 2500, -)
01:05:04.945 00.000 3013644336 MoveAxis(E, 2500, -)
01:05:11.547 00.000 3013644336 MoveAxis(E, 2500, -)
01:05:18.100 00.000 3013644336 MoveAxis(E, 2500, -)
For the DEC axis, North/South moves are problematic (22 North, 6 south):
01:05:24.634 00.000 3013644336 MoveAxis(N, 1250, -)
01:05:29.857 00.000 3015503888 Backlash: Rejected small move of 0.7 px
01:05:29.858 00.000 3013644336 MoveAxis(N, 1250, -)
01:05:29.886 00.028 3015503888 Backlash: Limpiando paso de backlash   2, Last Delta = 0.72 px, CumDistance = 0.72 px
01:05:35.109 00.001 3015503888 Backlash: Rejected small move of 0.8 px
01:05:35.109 00.000 3013644336 MoveAxis(N, 1250, -)
01:05:35.136 00.027 3015503888 Backlash: Limpiando paso de backlash   3, Last Delta = 0.83 px, CumDistance = 1.53 px
01:05:40.438 00.001 3015503888 Backlash: Rejected small move of 0.8 px
01:05:40.442 00.000 3013644336 MoveAxis(N, 1250, -)
01:05:40.470 00.028 3015503888 Backlash: Limpiando paso de backlash   4, Last Delta = 0.77 px, CumDistance = 2.08 px
01:05:45.789 00.000 3015503888 Backlash: Rejected small move of 0.6 px
01:05:45.790 00.000 3013644336 MoveAxis(N, 1250, -)
01:05:45.820 00.030 3015503888 Backlash: Limpiando paso de backlash   5, Last Delta = 0.58 px, CumDistance = 2.66 px
01:05:51.144 00.001 3015503888 Backlash: Rejected small move of 0.7 px
01:05:51.144 00.000 3013644336 MoveAxis(N, 1250, -)
01:05:51.169 00.025 3015503888 Backlash: Limpiando paso de backlash   6, Last Delta = 0.74 px, CumDistance = 3.32 px
01:05:56.380 00.001 3015503888 Backlash: Rejected small move of 0.8 px
01:05:56.380 00.000 3013644336 MoveAxis(N, 1250, -)
01:05:56.407 00.027 3015503888 Backlash: Limpiando paso de backlash   7, Last Delta = 0.84 px, CumDistance = 4.08 px
01:06:01.720 00.000 3015503888 Backlash: Rejected small move of 1.4 px
01:06:01.721 00.001 3013644336 MoveAxis(N, 1250, -)
01:06:01.750 00.029 3015503888 Backlash: Limpiando paso de backlash   8, Last Delta = 1.39 px, CumDistance = 5.46 px
01:06:07.013 00.000 3015503888 Backlash: Accepted clearing move of 4.5
01:06:07.014 00.000 3013644336 MoveAxis(N, 1250, -)
01:06:07.040 00.026 3015503888 Backlash: Limpiando paso de backlash   9, Last Delta = 4.49 px, CumDistance = 9.62 px
01:06:12.339 00.001 3015503888 Backlash: Rejected small move of 1.4 px
01:06:12.340 00.000 3013644336 MoveAxis(N, 1250, -)
01:06:12.365 00.025 3015503888 Backlash: Limpiando paso de backlash  10, Last Delta = 1.43 px, CumDistance = 11.05 px
01:06:17.627 00.000 3015503888 Backlash: Accepted clearing move of 2.3
01:06:17.628 00.000 3013644336 MoveAxis(N, 1250, -)
01:06:17.653 00.025 3015503888 Backlash: Limpiando paso de backlash  11, Last Delta = 2.28 px, CumDistance = 13.26 px
01:06:22.931 00.001 3015503888 Backlash: Rejected small move of 1.8 px
01:06:22.931 00.000 3013644336 MoveAxis(N, 1250, -)
01:06:22.959 00.028 3015503888 Backlash: Limpiando paso de backlash  12, Last Delta = 1.81 px, CumDistance = 15.07 px
01:06:28.229 00.000 3015503888 Backlash: Accepted clearing move of 3.0
01:06:28.229 00.000 3015503888 Backlash: Got 3 acceptable moves, using last move as step 1 of N calibration
01:06:28.229 00.000 3015503888 Backlash: North calibration moves starting at {1986.6,1533.0}, Offset = 18.7 px
01:06:28.229 00.000 3015503888 Backlash: Total distance moved = 21.2
01:06:28.229 00.000 3015503888 Backlash: Falling Through to state GO_NORTH
01:06:28.254 00.000 3013644336 MoveAxis(N, 1250, -)
01:06:33.642 00.000 3013644336 MoveAxis(N, 1250, -)
01:06:38.988 00.000 3013644336 MoveAxis(N, 1250, -)
01:06:44.318 00.000 3013644336 MoveAxis(N, 1250, -)
01:06:49.648 00.000 3013644336 MoveAxis(N, 1250, -)
01:06:54.989 00.000 3013644336 MoveAxis(N, 1250, -)
01:07:00.329 00.000 3013644336 MoveAxis(N, 1250, -)
01:07:05.650 00.000 3013644336 MoveAxis(N, 1250, -)
01:07:10.961 00.000 3013644336 MoveAxis(N, 1250, -)
01:07:16.247 00.000 3013644336 MoveAxis(N, 1250, -)
01:07:21.530 00.001 3015503888 NORTH calibration completes with angle=87.4 rate=1.822 parity=-1
01:07:21.563 00.000 3013644336 MoveAxis(S, 2500, -)
01:07:28.130 00.000 3013644336 MoveAxis(S, 2500, -)
01:07:34.708 00.000 3013644336 MoveAxis(S, 2500, -)
01:07:41.267 00.000 3013644336 MoveAxis(S, 2500, -)
01:07:47.803 00.000 3013644336 MoveAxis(S, 2500, -)
01:07:54.360 00.000 3013644336 MoveAxis(S, 1250, -)
01:07:59.615 00.001 3015503888 Omitted calibration alert: Advisory: Calibration successful but little south movement was measured, 
so guiding may be impaired.
Afterwards, there are many North guide pulses before the DEC axis starts to diverge, that does not occur at the first North guide pulse (even if they don't seem to be very effective). I will try to look what happens when the DEC starts diverging.
At first sight, I would think of a weight balance issue, or a loosen axis clutch. Or there may be a bug in the motor controller firmware as suggested by @Gonzothegreat, which the ASCOM driver is aware of ?
Maybe a log from an Ascom/PHD2 session would be helpful here. I'm also thinking to the SideofPier property, which seems to be used by some guiding software to use the correct axis direction.

Read More...