You can (and usually should) work with camera-native files to avoid quality loss, but the truth is MPEG-2 is so painfully slow, NOT because it is high quality but rather because of its processor-intensive format, you would be better off transcoding them to another format. This will help to avoid many other workflow problems as well, so speed is not the sole consideration here.
Most people consider MPEG-2/HDV to be an acquisition-only medium, and not a terribly great one either because it is the same datarate as mini-DV (3.1 megabytes per second, otherwise known as 25Mbps). The only way it is able to achieve HD frame sizes at a mini-DV datarate is by only using 2 actual frames for every 15 frames of video. The other 13 frames are just guessed by the codec and filled in with mush. It is all this “mush” that trips up your apps, since they don’t see actual frames.
The common workflow is to transcode the file to another type for post-production purposes. 8-bit uncompressed would probably result in no visible quality loss and would be relatively fast, but there aren’t a lot of good options available to you. You can’t really convert to another camera-native file type like DVCPRO HD without losing a lot of quality so it’s either uncompressed or accept some loss (perhaps with the motion_JPEG-A codec).
Ultimately the biggest problem is shooting in MPEG-2 because it limits your post workflow so badly, so you may consider buying a camera that shoots in a more standard format such as the HVX200 and its DVCPRO HD files.