I'd guess the EROEI is better than that for LNG, but I haven't seen any figures.

From Robert Rapier's article on XTL, with inconclusive chat in the comments :

Of course there is a catch. GTL is not all that efficient. There are efficiency losses during both the POX and the FT processes. It would be far more efficient to run automobiles directly on the natural gas. Due to the fact that the gas is stranded, this is obviously not an option. But the efficiency losses are significant. According to the Syntroleum link, it takes 10,000 cubic feet of gas to make 1 barrel of fuel. 10,000 cubic feet of natural gas contain roughly 10 million BTUs, but a barrel of fuel contains only around 5.5-6 million BTUs. Forty percent of the BTUs are either lost as radiant heat, or turned to steam and consumed in the GTL plant. Unless carbon sequestration is in place (unlikely), all of those BTUs wind up as carbon dioxide in the atmosphere. On top of that, the BTUs from the barrel of fuel are going to wind up as carbon dioxide in the atmosphere after the fuel is combusted.

How long this can compensate for oil depletion depends on how much gas were in turning into LNG and shipping offshore.