Either way, it seems to me that it is currently too easy for the network to get overwhelmed. We need a quicker solution than SegWit.
What's your proposal?
Then we can see if SegWit/Lightning can be deployed soon enough for adoption not to be hurt due to lack of scalability.
Edit: Just for clarity, I'll add that I don't believe SegWit alone to be enough.
Unfortunately, your beliefs do not correspond to reality.
Segwit puts a stop to 3rd parties malforming transactions. That means that 2nd layer solutions become possible, where now, they're not. Scaling on the first layer is not going to happen, it simply isn't possible.
The way to handle the 1st layer is to choose a blocksize that represents a compromise between number of network nodes and transaction rate. We're there, right now.
Nodes are at a local low point.
Transactions are at an all time high point.
That means 1st layer is exhausted, and that any attempts to use the 1st layer to increase the transaction rate can only end in the number of Bitcoin nodes decreasing. Have you not researched any of this stuff?
I believe the risk of not increasing the blocksize modestly (congestion) is greater than the risk of node count lowering (centralisation).
Do you believe 1MB is exactly right? If layer one is really that exhausted then surely we should lower it? I find it funny how so many people say we should not increase, but so few say we should decrease. 1MB was arbitrarily chosen after all.
Right now I see full blocks as a bigger problem than the number of nodes. 2MB or even 4MB gives us enough time for Lightning to be ready.
! 
