Remember this quote by Adam when the time comes that they make their proposal to change the Bitcoin protocol to enable LN and SCs.
As you can probably tell I think a unilateral fork without wide-scale consensus from the technical and business communities is a deeply inadvisable.
The bar to add LN and SC enabling protocol changes is wide-scale consensus. If a small number of participants disagree, it doesn't happen by their own logic. Additionally these protocol changes significantly change what Bitcoin fundamentally is much more than a simple 20MB blocksize c
This is the exact reason why I linked Adam's reply: quote for posterity the evaluation criteria used by opponents to gavin's block size increase proposal.
That said, if memory serves the new opcode needed by sc could be introduced by a soft-fork. Though I have to agree that merged mined side chains would introduce a significant modification of miners economic incentives.
One last thing, I think that at least Adam is not opposed to the increase per se, in fact he did say that LN and SC need a modification in the way max block size limit is handled (he talks about burst period), he even praise jgarzik bip 100 on the basis of the "slow" variation entangled in the proposal.
edit: fix typo