Coffee/alcohol and dehydration

An interesting question which is very hard to find scientific information about on the internetz is this:

How much alcohol does a drink have to contain to give a net loss of fluid (i.e. act dehydrating)?

Unfortunately, there seems to be no simple answer on this question. A couple of sites cite nutrition experts, such as this one, where a lecturer claims that alcoholic beverages containing more than 10% alcohol gives a net loss; however, the percentage seems to be taken out of thin air. Update: This old study from 1941 hints that on an average every 10 g of alcohol makes you pee out 100 ml, which gives a limit on about 9% (100 g fluid and 10 g alcohol = 10/110 = 9.1%). This would indicate that normal strength beer (~5% alcohol) will actually hydrate you, as only about half of the fluid gets lost due to the diuretic effect of alcohol. However, the response varies highly among individuals.

There are also a couple of random people suggesting that the response is highly individual and dependent on a lot of other circumstances. An anonymous comment here (last comment), for example, suggests that low strengh beer hydrates most people whereas a whiskey does not (duh!). This study from 1997 tells us that drinking 4% beer is worse than drinking water for rehydrating, however it is not clear from the abstract if it has a net positive effect. In this study, results indicate that the diuretic action of alcohol lessens when already dehydrated.

Another, related, question is:

Does coffee really act dehydrating?

This question seems simpler to find an answer to, and that answer is simply no. In fact, under normal circumstances and when not exaggerating intake, coffee can be regarded as water from a hydrating point of view (Wikipedia lists a number of sources). So that coffee acts dehydrating seems to be a plain myth.