[Piwoslaw]

From Wikipedia:

Tesla's design sidestepped the key drawbacks of the bladed turbine. It does suffer from other problems such as shear losses and flow restrictions. Some of Tesla turbine's advantages lie in relatively low flow rate applications or when small applications are called for. The disks need to be as thin as possible at the edges so as not to introduce turbulence as the fluid leaves the disks. This translates to needing to grow the number of disks as the flow rate increases. Maximum efficiency comes in this system when the inter-disk spacing approximates the thickness of the boundary layer, and since boundary layer thickness is dependent on viscosity and pressure, the claim that a single design can be used efficiently for a variety of fuels and fluids is incorrect. A Tesla turbine differs from a conventional turbine only in the mechanism used to transfer energy to the shaft. Various analyses show that the flow rate between the disks must be kept relatively low to maintain efficiency. Reportedly, the efficiency of the Tesla turbine goes down with increased load. Under light load, the spiral taken by the fluid moving from the intake to the exhaust is a tight spiral, undergoing many rotations. Under load, the number of rotations drops and the spiral becomes progressively shorter. This increases the shear losses and reduces the efficiency.

Efficiency is a function of power output. A light load makes for high efficiency and a heavy load, which increases the slip in the turbine, lowers the efficiency. This is not exclusive to Tesla turbines.

The turbine efficiency of the gas Tesla turbine is estimated to be above 60, reaching a maximum of 95 percent. Keep in mind that turbine efficiency is different from the cycle efficiency of the engine using the turbine.
[...]
In the 1950s, Warren Rice attempted to re-create Tesla's experiments, but he did not perform these early tests on a pump built strictly in line with the Tesla's patented design (it, among other things, was not a Tesla multiple staged turbine nor did it possess Tesla's nozzle). Rice's experimental single stage system used air as the working fluid. Rice's test turbines, as published in early reports, produced an overall measured efficiency of 36% to 41% for a single stage. Higher percentages would be expected if designed as originally proposed by Tesla.
[...]
Actual modern multiple stage bladed turbines typically reach 60% - 70% efficiency. Actual volute-rotor matched Tesla-type machines of reasonable size with common fluids (steam, gas, and water) would also be expected to be around this range (if not higher).