I started a similar thread entitled
Block Spacing And Security a while back, as I'm very interested in this as well. Specifically, whether 10 minute block intervals are necessary. (Of course, a hard fork like lowering block spacing will likely never occur with Bitcoin, which creates an interesting case of path dependence.)
tl;dr Security is different than decentralization. Security does not depend on block spacing.
Security is total amount of work done to secure a transaction from blockchain reorganization. Say a bitcoin block has X amount of work. For someone who receives a transaction and waits 2 blocks you can say his funds are secured by 2X amount. If the receiver waits 20 blocks he can say that the received amount is secured by 20X amount.
Now since litecoin has 30 less value and 4 times more blocks to secure the received transaction that is equivalent to 1 bitcoin block you have to wait 120 litecoin blocks. So bitcoin is much more secure. (Note that proof of stake has precicely 0 security because one or more past stakeholders can use their stake and reorganize the chain from the point before selling at 0 cost)
What I am asking here is different and refers to decentralization. (Even now that ghash controls a lot of hash they are contributing to the security as long as they don't use to create a parallel chain and thus stop solving main chain blocks)
What I am speculating is if spacing helps pool decentralization and does not harm the decentralization as claimed by some devs (I think gmaxwell made such claim in the past). Just wanted to ask why this might wrong or has some premise. Definitely popular alts like doge make a good study case