All technobabble. Usually, if you don't know what it is or what will change, leave it and stick to the profile or look for user based guides on which values are worth changing. Very often there is need to burn a trail in the video industry!

Many compression techniques start off with a mathematical transform called the Disrete Cosine Transform, or DCT. The parameters of this transform and the steps that dictate the block size, co-efficients, number of pictures of each type (IBP) in a group all equate to various degrees of complexity in your encoder/decoder.

Usually, if you know what your target format can accept then you make adjustments based on this or stick to the Sony templates. One good reference for handling the various MPEG-2 elements is from the manufacturer info and guides on how to configure Tsunami MPEG Encoder, TMPGEnc 2.5.x.x from Pegasys-Inc. I've come across many expert tuners of MPEG-2 encoders who swear by using TMPGEnc over the Sony provided MainConcept encoder, usually via the means made possible from the Vegas-project/timeline using debugmode's frameserver (saving a double render via a low compression codec). For anything over an hour on a single layer DVD-Video DVDR I tend to look towards TMPGEnc and I get fairly predictable results from the guides I've collected with regard to tuning and pre-filtering in that encoder.

Main Level is an MPEG-2 level name (and is usually quoted also with a named profile)
http://en.wikipedia.org/wiki/MPEG-2

DC co-efficients, tuning for use by the DCT:
quoting from MPEG.org
"For each category enough additional bits are appended to the SIZE code to uniquely identify which difference in that category actually occurred. The additional bits thus define the signed amplitude of the difference data. The number of additional bits (sign included) is equal to the SIZE value."

I read into this that the values in the compressed data have so much apparent dynamic range but this can be enhanced by a scaling factor, ie the co-efficient where necessary. So if your target decoder (ie a DVB or DVD-Video device built to adhere to the standard and no better or worse) can support something other than the Vegas template's default that you are copying from then you might see some visual quality advantages. There isn't enough information available about the MainConcept codec to know if changes to complexity will also increase your encoding time or if the higher values will help ALL types of scene or clip.

I'd imagine "I frame and b frame 1-7" would refer to the number of B frames to be provided after each I frame. B frames are like P frames in that they contain picture 'differences' or deltas, but serve to aid smoothness when the viewer is using this MPEG-2 content in a media format where he wishes to reverse play or reverse-wind with picture. [edited here>*] Although as pointed out by a learned friend of mine (thanks ScorpioProd!), the emphasis is on the compression efficiency that a calculated 'B' frame enjoys from looking both forward and backward to make the most suited delta bi-directionally. The benefits for reverse play are the subject of specific /arbitrary implementations and not the compression design itself.

Write-sequence end-code is MPEG-2 punctuation for the end of the program.
http://download.sonycreativesoftware.com/whitepapers/mpeg_overview.pdf
By having it disabled or empty you can then splice streams such as these together if you have the right tools and are not breaking any rules for the format if the next file/stream has very different parameters.

I hope the links are entirely useful.