I have yet to see any good evidence in support of this. The best controlled experiment that I've seen was reported many years ago on the Usenet group rec.bicycles. The chain was split in half and reconnected using two quick-links and a tiny scratch marked on one of the links to distinguish the two halves. Every 500 miles of so the chain was removed, split in two, and only one half was thoroughly cleaned. The two halves were then measured, reinstalled, and relubricated. After a few thousand miles a definite trend was observed in the measured lengths with the half that was repeatedly cleaned consistently measuring longer (i.e. more worn) than the half that was never cleaned.
While I don't remember the details of the cleaning procedure, I do recall that it seemed quite thorough and that care was taken to ensure complete lubrication of the interior of the chain afterwards.

I don't normally do any chain cleaning and wouldn't consider doing it on tour. Finding a replacement chain every 10000 km or so doesn't seem like too much of an inconvenience.