Will predictions about the diappearance of religion ever disappear?

I imagine the BBC must have time on its hands.  How else can one explain its decision to post, last Tuesday, yet another article on the question of whether religion is about to disappear?  The BBC's reporter, Rachel Nuwer, attempts to make the issue appear new by citing a recent Gallup International Survey showing a decline in the number of believers over the past few years.  Worldwide, Gallup claims, the number of people claiming to be "religious" declined from 77 percent in 2005 to 68 percent in 2011, while the number of avowed atheists has increased from 10 to 13 percent.

Nuwer and the BBC do deserve some credit for not coming to the conclusion that religion's demise is imminent.  But the Gallup surveys on which they rely for their claim that people have become less religious all around the world in recent years are problematic.  Gallup subcontracted much of the actual surveying work out--which makes one wonder how good a job subcontractors in each country did of getting representative samples of the population. 

There is also the question of method of contact.  Gallup notes that the survey was conducted using a mix of face-to-face, telephone, and e-mail interviews--but only one method was used in each country.  Given the decline of the landline over the past decade, and the concern it had caused the CDC in gathering information on Americans' health, one wonders whether surveys conducted through only one mode of contact can any longer be reliable.  The Economist notes that in United States, homes without landlines tend to be disproportionately young, poor, and Hispanic; the United States Census has noted that homes without internet access are disproportionately poor.  The latter fact alone would call into question the survey's results concerning Americans' religiosity.

Then there is the whole question of what it means to be "religious" in the first place.  The word does not have an easy equivalent in all languages; even Biblical Hebrew itself contains no word that would exactly translate as religious, describing characters only as people who "fear the LORD".  One wonders if, say, Saudi and Irish respondents to this type of survey are operating with the same definition of what it means to be a "religious" person.  Add to this the influence religious repressive governments--whether officially Islamist governments such as Oman or officially atheistic ones such as China--can have on people's willingness to self-report their own religiosity, and it becomes even harder to trust such a survey.

But more to the point, Nuwer never mentions how often the question of religion's eventual disappearance has been raised before.  Enlightenment thinkers such as Voltaire predicted the demise of religion; in the 1960s, noted sociologist Peter Berger made a prediction that religion would be gone by the year 2000, a position he was later forced to admit was premature.

A little research might have stood Nuwer and the BBC in good stead.  But that assumes their purpose was to analyze recent events in the news rather than simply to be sensationalistic.