Saying that liquidity reduces is a fallacy that was believed quite heavily pre-2007.
Liquidity does not reduce risk, it has no effect on risk because it has no effect on outcomes. If you invest in a stock that goes bankrupt, there is no way to get out of that decision with no cost.
Equally, someone is only willing to supply that liquidity if they get something in return. So in those worst cases, you will usually not have the liquidity that you think because someone needs a return for taking the elevated risk in those situations (and again, that risk doesn't go anywhere, liquidity just shifts that risk to someone else).
Look at the market, is there really a deficit of risk-taking? I think there is a deficit of understanding of the risks but there is a massive excess of risk-taking in pretty much every asset class.
I also wouldn't associate illiquidity with return either, that is quite wrong. The decentralized market-making model that we have minimises costs and maximises liquidity but this is just a net gain for investors (due to the clear demarcation of types of risk-taking within the system). There is no real illiquidity premium, and the societal gain from higher liquidity is just a pure gain for the system (again, I am not saying it reduces risk, it does not but it does reduce costs, increase transparency, and produces a system that is robust relative to bank-led financing systems common in Europe, for example).