You are right to be suspicious. So let me be a bit more nuanced and guarded and explain my claims.
I claimed two things.
First, the central bank can always increase aggregate demand. My argument follows standard market monetarist lines.
Demand is just another word for money that's being spent. The central banks can increase that amount by `printing money'. In practice, this has meant buying government bonds, but the central bank could buy all kinds of assets (like real estate or stocks). http://www.wsj.com/articles/new-tool-for-central-banks-buyin... talks about the Bank of England buying corporate bonds. Or they just drop the newly printed money from helicopters. (https://en.wikipedia.org/wiki/Helicopter_money)
(And as a second prong to my argument: if buying assets would not increase inflation, the Bank of England could just buy up eg all of America, and give each Brit a nice big ranch. And if helicopter money would not increase inflation, the Bank of England could just infinitely finance a lavish basic income for everyone directly.)
If any central bank in charge of a fiat currency ever seems to have problems creating arbitrary large amounts of demand, that's purely a political issue holding them back, not lack of ability. (I explicitly say `in charge of a fiat currency'. Eg a gold standard makes it much harder to `print money'.)
Of course, ideally we the central bank would create enough steady demand to avoid layoffs and economic slumps, but not drive inflation above something innocuous like a 2% target. We don't want Zimbabwe style inflation. Instead of targeting the growth of the price level (= inflation), a better approach is to target the trajectory of the price level over time directly; or to target the trajectory of nominal aggregate income. For historical reasons the latter approach is called nominal GDP level targeting. (https://en.wikipedia.org/wiki/Nominal_income_target)
Second, I claimed that land value would be a good base for taxation. My argument follows standard Georgist lines, and is basically mainstream economic orthodoxy. (See eg https://en.wikipedia.org/wiki/Land_value_tax and http://www.henrygeorge.org/rem42.htm)
That claim has two parts: (a) there's enough land value to be taxed to create significant revenue, (b) taxing land value would not disturb the economy.
What do we mean by a land value tax (lvt)? An lvt is a recurring tax on the unimproved value of land. `Unimproved' means that we subtract the value of any structure built on top. The aim is to tax land as a factor of production, not capital or labour.
Actually, we want to be taxing the recurring income that accrues from land. Whether income ever hits the market or not, as in the case of an owner-occupier. The easiest way to do that is to tax a proxy, the land value.
When you tax anything, you usually get less of it. That's welcome in the case of eg alcohol consumption or sulphur emissions, but not in the case of working or investing.
Mark Twain famously quipped "Buy land, they're not making it anymore." That means, that no matter how we tax or subsidies land, we are going to have the same amount of it in a specific territory.
The landlord also won't be able to pass on the tax to tenants or leasing companies. (See https://en.wikipedia.org/wiki/Tax_incidence for why that is so.)
There is already a lot of land rent to go around, but when we decrease other taxes, the land rent increases. (You can see that empirically in the areas around eg the Swiss/German border: the Swiss side has lower taxes and higher land values.)
The idea is to shift all (or almost all) taxation from labour and capital to land.
Anything questionable about these things?