(more users means more nodes and a higher bitcoin price which means more miners)
Right, well that's wrong.
More users won't and can't amount to an increase in nodes; if increasing the blocksize prevents them running a node, they'll end up using web-nodes or non-verifying wallets instead (i.e. reduced/zero censorship resistance).
Last year, I spent time staying in middle-class rural area, in a Western country.
Maximum internet speeds were 0.5 MB/s dl, 0.5 MD/s ul. Only last year.
When I tried keeping my full node online in that setting, no-one else could use the same 0.5MB/s internet line for anything except baaaaaaasic webpages. If the webpage had more than a few megabytes of images, it was either too frustratingly slow to load, or just failed outright.
If that's what people in rural Western countries can expect, those in urban 3rd world countries won't be getting alot better. We need those people. So a conservative approach is best for achieving both censorship resistance
and decentralisation with this type of system. Decentralisation is a vital component to censorship resistance when you factor in the resources required to use Bitcoin in a censorship resistant fashion.