You have a major misunderstanding of what you've quoted. I honestly cannot find any similarity between scrypt and Xn algos (and there's no such statement in your quote) but if there is, by the same metric we can state safely all algorithms are the same.

X11 has been very successful as far as I am concerned and it's still far from being "failed"; I'd be interested in hearing OP argumentation in that sense.

I'd agree long chain algos seem to be getting less attention lately. This is possibly because of at least a few reasons:

• Besides Dash, nothing running long chained hashing seemed to be successful;
• There are persistent rumors of X11 FPGA miners. Those rumors appear unsubstantiated to me but others can have their own opinion;
• People insisting on putting GPU out of the picture (aka: let's just pretend 20 years of evolution in computing architectures never happened);
• This "memory hard" thing. While being "memory hard" is a necessary requirement for ASIC resistance, Xn algos have no memory hardness at all so they get the boot;
• From a purely theoretical point of view, chained hashing is nonsense (the hash of an hash is just as strong as the first hash), but I bet this is basically irrelevant in making it less trendy;
• New algos are new so they are BETTAR!!!!eleven! What's wrong with you! We are obviously making progress here! It's not rocket science, just applied cryptography, even my dog could figure it out in the time of a walk in the park while being trained in playing the lyra!
• ... ?