I think they are essentially against the idea that the block limit of 1 MB should remain.  Segwit as such they don't care too much about.  I think the question is rather in the other direction: why are segwit supporters (Core essentially) AGAINST block size increase ?  The two main arguments they advance for this:
- only soft forks because hard forks could lead to a chain split
- there should be many full nodes to keep bitcoin decentralized, and bigger blocks will kill full nodes

are bogus.

They are bogus because a well-done HF is much less of a danger to split the chain, than an aggressive, non-consensual soft fork, and UASF is the prime example of this ; and they are bogus because first of all non-mining full nodes have no consensus decision power, but moreover, the trick with the witness data is de facto increasing the amount of data that the full checking nodes would need to vehicle, so while a few years ago, "increasing the data flow" to 2 MB was out of the question because of the poor full nodes, now, 4 MB is not a problem if it are "witness data".

We clearly see that in all of this, there is only one single agenda: pushing people out of legacy transactions because of artificial block chain restrictions.

And all of this has nothing to do with "decentralization" (that is, many non-colluding entities having individual decision power): the decisions in bitcoin are in any case taken by a very small oligarchy of mining pools.  That's the bare reality of the state of decentralization in bitcoin.  

Litecoin was different, because 1) litecoin can be seen as a toy/testing bed so that people wanting to play with segwit can have their coin to play with ; there was no desire in Litecoin to keep small blocks for ever by Lee.  It would have been silly to require larger blocks when the current blocks are only filled at the percentage level.