As far as I understand, interlaced frames aren't complete frames, I think every other line of pixel is recorded and then in the next frame the missing lines - 1, 3, 5 first and then 2, 4, 6 second. The files run at a faster frame rate with potentially less blurring and it doesn't affect the filesize.
Deinterlacing I guess merges the extra frames so you get complete frames.
If interlacing reduces blur, if you deinterlace it you don't actually change the picture so the video should be just as crisp, just more compatible.