Well this may not be exactly what you need but I’ll share my experience with you. If you create a portion of video or an animation that will later be used and recompressed in a video, I suggest that you jack up the resolution to multiples of the finished product. So if you are making a finished movie of 320 X 240, work the animation in a multiple, like twice the resolution, 640 X 480. That way, when it gets recompressed, there are some pixels to spare.
You say your dimensions are 400 X 320, is that intended to cover the entire screen? The reason I ask is because that is not a 4:3 ratio, which is what all standard video is. Most video from digital cameras will also have a frame rate of 29.97 frames per second, not 28. Part of your problem could be coming from the distortion of a 400 X 320 squished into a 4:3 aspect ratio. If the video is 400 pixels wide, it should be 300 pixels high (4:3). If it is 320 pixels high, it should be …. ahhh … let me think.. It should be 426.66664 pixels wide. So that may be part of your problem.
The other comment I have is about text. Images of text are rarely as sharp as text added as titles, etc, during the last edit. That’s because the image of the text is being compressed during the rendering process of movie making.
Also be sure that your bitrate is as high as possible during all preliminary development stages (creating the animation, etc.). Again, use a multiple (like double) of the final bitrate. So for example, if your final .flv will be 400kbps, anything you put in and prepare for the transcoding should be at least 800kbps bitrate.
Best of luck to you on your project, let us know how it is resolved.