Trouble when computing A006743

[N. J. A. Sloane]

About a week ago there were several postings about trouble
with the recurrences for A006743.  The current entry gives
three recurrences stated by Simon Plouffe.  I have labeled
all three "May be incorrect, needs checking".  Can anyone
tell me me more about them?  Are /any/ of them correct,
or should they all be deleted?
Thanks
Neil Sloane

(Please wait 10 minutes before checking - I am just updating
the OEIS this minute)