You can download the 230 games played in the 1929-Karlsbad tournament
But how does that help?
That file has a very late source date (2003.01.01), and as it does not
reference any earlier works, it must be judged on that date.
The dates of the sources used by the original poster is, at a
reasonable guess, earlier than so. So how do we rule out the possibility
that one of the sources we're trying to decide the trustworthyness of
was used to create the chesslib archive?
Simple -- we can't. And so we can't rely on chesslib for this particular
question, unless there's something else to base such a decision on.
Source criticism is a must for this type of question. And for
that source references are needed.
Anders Thulin http://www.algonet.se/~ath