Neandertal Place in Human Origins

I think I might be a bit behind the times here, or ahead, I don't know which. I thought that recent research had decided that modern humans didn't descend from Neandertals, but had descended from Cro-Magnons. For example, Cheddar man has Cro-Magnon mtDNA, not Neandertal. I'm going on the work of Prof Brian Sykes and his mtDNA analyses of modern and ancient man. Furthermore the evidence seems to point to Neandertals and Cro-Magnons existing in the same time period 34-36000 years ago with the question of whether they ever interbred. The bottom line was that basically Neandertals died out as Cro-Magnons became more numerous. The mtDNA analysis suggested that Cro-Magnons came out of Africa in the same way that Neandertals did, but later and although they ovelapped a bit, they were distinct. Also a find last summer in Portugal has thrown up the possibility that Neandertals and Cro-Magnon may occassionally bred. The find is that of a child's skeleton and is dated to 34-36000years ago, during the overlap and certain features seem to be Neandertal and certain others seem to be Cro-MagnonNow, have I missed something that's happened to totally contradict this Cro-Magnon theory and completely invalidate the mtDNA analyses on which this theory is based? Please help, because I'm getting really confused!!!

Thank you for your reply. Must admit I was worried for a minute. I'd just got my head round the whole Cro-Magnon business, then I see the above talking about descent from Neandertals.I would be interested in any info anyone has on criticism of the mtDNA methodology used. I know at one point that evidence was put forward suggesting that recombination may have occurred ie that somehow mtDNA from the father's sperm could have gained access to the innards of the egg and thus changes in the mtDNA would not necessarily represent a mutation, but a huge recombination event, thus destroying the "timeline" used to determinehow long base changes took to occur randomly, without external input. I also know that 18 months after this was published, the authors retracted, having misread the data coming off their automated sequencer - they had claimed that they had found mutations at a position in the seventies of the sequence which proved recombination, but actually it was a mutation in the eighties which was already accounted for.I'm really interested in this and have been busily downloading primer sequences for the amplification of mtDNA and known sequences of the 500bpr fragment that they use in their analysis (in the vain hope that I might just find the time to do my own mtDNA). I would hate to think that all my hard graft during the Christmas holidays had been a waste of time. I really must get out more!!