Unable to increase playback speed.

Ok, so I'm using the latest codec pack as of August 5th 2014 (10.6.4 Mega) LAV and AC3Filter for everything, ffdshow for raw video, MadVR renderer, Windows 7 Ultimate, Nvidia CUVID...

It is currently impossible to have ANY file type playback at higher than 1.00 rate in MPC-HC.

Any way to fix this?

Increasing playback rate only works if all filters in the chain support it. I have tried with the same set of filters and increasing the rate working fine. However, I can imagine that it fails because of a performance bottleneck.

Some things you can try:
1) DXVA2 copy-back instead of CUVID, or even software decoding
2) different video renderer, or less heavy settings in madVR
3) disable processing filters in ffdshow

I've tried everything except changing away from CUVID, including changing renderers...

Just tested all other decoding acceleration techniques... No speed increase ability. (and CUVID has nice hardware-based deinterlacing, so I'm sticking with that when possible)

This officially makes it to where I have tried all possible settings and settings combinations of that renderer and splitter combination, including on the newest updated codec pack. (10.6.5)

Just in case it is a hardware problem: I'm running everything on a Lenovo G780 laptop with a GT630m 2GB dedicated video card, an i5-3210m processor, and 8GB of 1600 RAM with low latency.

If your objective is to "fast forward", then another method would be to use the jump function. Ctrl+Left/Right does a 5 second jump.

Yes, I do that already, but I also have some video that I wanted to play back at 1.25x-2.5x speed. The goal is not to jump to a later time, but to be able to fully observe the video in it's entirety at a faster rate. (for example watching 80 hours of security camera video in real time looking for a certain individual that shows up for 10 seconds is impossible to do without it being a full-time job)

I understand.

I have no idea why it doesn't work on your system. You could try with VLC player instead for this specific use case.

I'll give it a go, thanks for trying.