I've just read the whole thread, as well as the report and the newspaper article.
Being pg and unmarried it has been quite useful as we had just recently been wondering if we should perhaps get married for legal reasons but didn't know the implications. The commitment aspect of marriage is totally unimportant to us as the decision to try for a child together is, for us, the strongest form of commitment we could possibly imagine. Anyway, many thanks to those who took the time to explain.
I'm glad the thread has clearly demonstrated that the message attributed to the figures is clearly, undisputably, false.
I would like to add that it is not just the DM that misinterprets the figures, but that actually the report itself contains strong implications, suggesting (but not actually saying) the same as is found in the DM article.
One major point which could perhaps still do with some clarification: The question of selection or causality. Many here have argued that getting married will not change the likelihood of splitting. Those couples that work will most likely work with or without marriage certificate, those that don't work won't work either way. That is an excellent point and one which cannot be emphasized enough. It is the same as saying that although there is a correlation between marriage and relationship stability (which the figures seem to imply), that doesn't mean that there is causality, i.e. that one causes the other. The alternative explanation for the correlation is the selection argument: If married couples are statistically more stable than unmarried couples, then this can just as well be down to stable couples being more likely to get married than unstable couples. Ergo, marriage does not cause stability, but rather, stability means people select marriage (on average). Choose for yourself which sounds more plausible.
Now the report starts out admitting that nothing can be said about this question - selection or causality - based on the current data. That is probably one of the few really "true" and un-ambivalently correct statements in the report. The author does go on to immediately belittle the "selection" argument rhetorically. And also goes on to imply that marriage indeed causes relationship stability (although he starts by saying that it is impossible to make such statements).
So. This report says nothing at all (but incorrectly implies a lot) about the question if marriage helps relationships or not. And yet it is used as an argument for promoting marriage for the sake of the kids who suffer from relationship breakdown. Go figure.