|Journal of Integer Sequences, Vol. 13 (2010), Article 10.1.4|
In this paper the authors ask if it is always true that g0 < g1. (To make it interesting, we need to rule out trivial cases where one number is in the N-span of the other elements.) I found many counterexamples, of which the smallest seems to be (8,9,11,14,15), with g0 = 21 and g1 = 20. (I do not know if there exists a 4-element counterexample, but I doubt it.)
Received January 11 2010.
Addendum, September 7 2010: J. Stankewicz found the 4-element counterexample (28,30,105,168) with g0 = 467 and g1 = 425. The smallest is (10, 15, 32, 48) with g0 = 101 and g1 = 99.