If there turns out to be some fundamental flaw in SHA256 (which seems very unlikely), then bitcoin could switch to some other algorithm.
Litecoin has helped finance the development of ASICs for "some other algorithm", thus makes the idea of being able to switch algorithm look more plausible than would be the case were there no alternative algorithms for which ASICs are available.
Thus to the extent that litecoin continues to motivate improvement in alternative-algorithm ASIC technology and availability, it is potentially useful to the Bitcoin ecosystem.
Of course if SHA256 does not turn out to have some serious vulnerability/flaw, all that work could turn out not to have been needed after all.
If someone makes ASICs for a more efficient alternative, then using scrypt as fallback algorithm could easily come to be seen as a poor choice.
It is ironic, of course, that it was propaganda claiming to aim at "ASIC resistance" that has financed the development of a fallback-algorithm line of ASICs.
The drawback to using such irony to develop a fallback is the expense and/or inefficiency of the fallback line of ASICs.
-MarkM-