There are rarely 1 conventional reactors on a site, although there are a few examples to the contrary. The largest sites have 6-8 reactors. Typically there are 2-4, which facilitates load management during outages.
The good news from the security prospective is that SMRs will likely be similarly concentrated, just with greater numbers of reactors. They are highly unlikely to be sprinkled about willy nilly, despite the thinking of some. Siting is a major problem with all power generation and particularly with nuclear: getting a site approved is a major accomplishment involving no end of costly wheeling and dealing over many years.
Although there are use cases, often championed as killer applications for SMR, such as very remote, undeveloped locations, I expect that for every 1 such isolated reactor actually deployed, 100 will be lined up in rows and columns at major reactor sites.
A big consideration is grid structure, it's very expensive to create power distribution from generators into the infrastructure. There already exists grid infrastructure at old reactor sites, so I think it's very likely that they would be ideal candidates for SMR deployment.
Some fraction of existing coal and LNG sites, and their existing infrastructure, are also feasible. Larger sites are often quite distant from populations and have large perimeters.