In the vast majority of (if not all) languages that have a grammatical gender assignment with the Masculine/Feminine distinction, the masculine is the unmarked gender. But the system of markedness is going to be much more complex as genders proliferate. For instance, in Czech verbal agreement masculine is the unmarked gender with respect to feminine and neuter but feminine is the default with respect to neuter. Also, the masculine inanimate is marked with respect to masculine animate (at least where agreement is concerned). As Corbett pointed out in Gender: markedness is not a very good category to apply across all of gender across all language.
It is important to keep in mind that the vast majority of languages, including English, do not have grammatical gender - http://wals.info/feature/32A#2/25.5/148.2. In English, the gender is lexical or semantic (e.g. the distinction between he/she, actor/actress) in contrast to a language like Finnish which does not have even these distinctions.
However, there is also the cultural default, which in most contexts skews strongly towards the masculine. E.g. the cultural expectation that a doctor or boss is a man. This is salient even in cultures where grammatical gender exists and most statements are gender marked. It is impossible to perfectly disagregate the role of grammatical markedness from that of the cultural semantic default, for instance in 'los ricos', grammar plays a role but it is also guided by culture. Another example is the English he/she debate where clearly the supposed grammaticality of the he-default did not withstand a cultural challenge (and was indeed itself a result of an earlier cultural challenge from the other direction).