Expensive prime arable land in the UK is rentable for £200 per hectare per year. You can buy it for £20k a hectare (land value is far higher than usable value due to tax avoidance schemes -- nobody spends £20k on a field to rent it out for £200 a year)
But even if you buy 100 hectares of land for £2m, you then need to build your 60MW, which costs about £60m on top and generates 60GWh a year or £6m at 10p/unit
The land cost doesn't matter.
The point is the electric delivery problem is much cheaper if the power generation and the power delivery both happen within the same substation (eg in the same neighborhood). If the generation happens in rural Scotland but the consumption happens in London, then all of the transmission needs to be upgraded. The point of your parent's comment was to describe a trick to minimize additional transmission costs/development.
And London area land does cost, even if it is just borrowing a vacant building's rooftop.
I don't necessarily agree with your parent's comment, but I don't agree with your comment either.
You don't get a benefit of reducing peak demand from on-building solar, so you can't do with smaller substations.
The legacy grid of producing power far out makes no sense. Coal, nuclear, hydro were far away and needed transmission. New grid can produce energy in the cities. Think of all the surface area available on all the buildings. Just the walls/fences between buildings is sufficient to build massive amounts of vertical solar.