I am almost certain that the bitrate is not distributed evenly among them.

I tought that a 5.1 standard 448kbps was made of 4ch of 96kbps each + C at 64kbps + a very small amount for the optional LFE, all indipendent encoded ...

Like the famouse mp3 CBR 128kbps is made of 2ch of 64kbps each.

The actual AC-3 encoder is more complex than indicated in Figure 1.2. The following
functions not shown above are also included:
1. A frame header is attached which contains information (bit-rate, sample rate, number of
  encoded channels, etc.) required to synchronize to and decode the encoded bit stream.
2. Error detection codes are inserted in order to allow the decoder to verify that a received frame
  of data is error free.
3. The analysis filterbank spectral resolution may be dynamically altered so as to better match
  the time/frequency characteristic of each audio block.
4. The spectral envelope may be encoded with variable time/frequency resolution.
5. A more complex bit allocation may be performed, and parameters of the core bit allocation
  routine modified so as to produce a more optimum bit allocation.
6. The channels may be coupled together at high frequencies in order to achieve higher coding
  gain for operation at lower bit-rates.
7. In the two-channel mode, a rematrixing process may be selectively performed in order to
  provide additional coding gain, and to allow improved results to be obtained in the event that
  the two-channel signal is decoded with a matrix surround decoder.
.

Few people even know that an mp3 CBR 128 is L 64 kbps + R 64 kbps!

* -m s/j/f/d/m    stereo mode
Joint-stereo is the default mode for input files featuring two channels..

stereo
In this mode, the encoder makes no use of potentially existing correlations between the two input channels. It can, however, negotiate the bit demand between both channel, i.e. give one channel more bits if the other contains silence or needs less bits because of a lower complexity.

joint stereo
In this mode, the encoder will make use of correlation between both channels. The signal will be matrixed into a sum ("mid"), computed by L+R, and difference ("side") signal, computed by L-R, and more bits are allocated to the mid channel.

...
...

dual channels
In this mode, the 2 channels will be totally independently encoded. Each channel will have exactly half of the bitrate. This mode is designed for applications like dual languages encoding (ex: English in one channel and French in the other). Using this encoding mode for regular stereo files will result in a lower quality encoding.

...

Is there a way to lossless extract just the L + R channel from an AC3 6ch to a new AC3 2ch?

There's a difference between extracting L & R and downmixing to Stereo.

Few people even know that an mp3 CBR 128 is L 64 kbps + R 64 kbps!

Because it's clearly not. Only a very crude encoder would do this.

Quote from: SebastianG on 2008-05-13 16:58:59

There's a difference between extracting L & R and downmixing to Stereo.

Yes: lossless operation!

And what if your L or R channels were coupled  together with some of the channels you wish to drop? You would have to go quite deep into the decoding phase - up to channel decoupling - and re-encode from this point.