Excuse me if I understand something wrong... I'm a girl who dabbles in Java, and some of it might be over my head, but if this were to be implemented, a "hard 51" would go like this:

1) attacker uses Superior ASICs to mine a private BTC chain forked from some point long since past (probably so as to make a double-spend)

2) attacker publishes chain

3) an unspecified mechanism detects that there is, well, a so-called "malicious reorg", and BTC enters "hide yo coins, hide yo wife mode" and starts relying on backup data embedded in Litecoin's chain

It seems to me that

a) it would be exceedingly hard to properly develop detection algorithms for "malicious reorg" detector

b) if you succeed at a) you don't need no litecoin, you just keep around the blocks being reorged away until you are convinced that reorg was "not malicious".

c) you can, of course, just halt everything and have community "heavy hitters" decide in IRC which part of the reorg was "good" instead of doing it with "and then a miracle occurs" automagical "malicious reorg" detector, but as long as you don't discard the blocks involved (which does not require litecoin) you don't really need litecoin in this process.



So, TLDR:
I do see theoretical benefit to this general line of thinking (the "let's detect bad reorgs" line of thinking

I doubt that "malicious" reorg detector that would activate a "hide yo coins" mode when a real attack occurs is attainable

I could see a "dumb" detector being used to facilitate a "consensus intervention" on part of humans who have nontrivial power over the network (pool ops and devs).

I do not quite see why you need litecoin in any such scheme, beyond "let's give this altcoin some semblance of purpose"