I did run into a test run by an engine builder's magazine where they strapped an inexpensive EBay turbo to a GM 355 V8 which had 150k miles on it. They dialed it up to around 800bhp with 8psi or so of boost. So, if plumbed in properly the cheap turbos do function, but they were not tested for durability. Repeated heating and cooling expands and contracts the cast iron exhaust impeller housing. It has to be very good cast iron, something that can withstand extreme heat ranges and pressures without cracking or exploding outright. The aluminum intake compressor will also expand and contract, but at different rates than the impeller housing. It has to withstand the high internal pressures of a compressor without blowing up. When these expand, the housing to turbine blade clearance shrinks, potentially causing the blades to contact the housing. This is unlikely, but in cheaper turbines where tolerances are guessed at, the gap must be overcompensated for. So, they can either be less efficient, or more prone to blowing themselves to pieces. Another consideration is the quality and tolerance of the floating bearing. The center rotating hub must float on a high pressure, externally cooled oil supply. This is almost always supplied by the engine's lubrication system, so it must be plumbed in and out of the existing oil loop. The quality and spec of the oil and the tolerances withing the bearing and center rotating hub are critical. As the steel shaft heats up, it closes the gap to the bearing. This clearance must also be accounted for at a range of temperatures. Some of the turbo kits just use a t-splice to take pressurized oil from the oil pressure sending unit line. I'm not sure if that's a good solution, as the flow is pretty restricted in such a small line. I wouldn't throw one of these turbos on an engine without adding an external oil cooler to prevent cooking off the engine oil. I would probably plumb the turbo directly into the cooling unit circuit. Most after-market oil coolers use an adapter plate at the oil filter connection to send off oil to a heat exchanger. This is not a bad idea in general, and requires no significant engine modification or machining. Expect engine oil to break down much quicker when run through a turbine. Water cooled turbines are much more expensive.
So, the turbo itself may look cheap, but to make it work properly you need the following:
-oil plumbing: $100
-oil cooler and plumbing: $150
-intercooler and air intake plumbing: $150
So now we are at $670. That's still cheaper than one Garrett turbine without all of the plumbing. However, there are more problems which must be accounted for. What happens when you add a whole lot more air to an engine which wasn't designed for a lot of boost? The fuel/air mixture is now very lean. Lean mixtures are bad news. As you run a lean engine up the rpm range, things start to get really hot. There isn't enough cold, high octane gas to cool the cylinder down. Fuel starts detonating before the piston reaches its apex. This is where engines burn up valves, blow holes in pistons and in the worst case throw rods through the block. The maximum fuel pressure and injector capacity isn't enough for the additional air; it can't keep the air/fuel mixture rich enough. A rich mixture prevents detonation to some degree, as does high octane fuel or possibly adding an ethanol mix. Some engines will run fine at lower RPMs, but as the turbine spools up and the engine starts breathing a whole lot more, the injectors run out of capacity. The easiest solution is to run higher capacity injectors and add a high pressure/high flow fuel pump. Electronically controlled engines with detonation sensors, MAF, O2, and EGT sensors can alter ignition timing and fuel mixture to get optimum fuel burn without detonation. Older engines without a feedback loop may use air flow to regulate fuel flow similar to a carburetor, but using a pressure regulator rather than the Venturi effect/float system. So the fuel system must be altered to either run richer overall, and not lean out at the higher RPMs, or run a mixture regulator manually which can be adjusted manually. So, considering the fuel system alterations, the cost will be well over $1200, which is still reasonable for turning a boring 180bhp into 400 or so clutch smoking horses.
And then back to the original reason for this post, will the EBay compressor hold up for a reasonable amount of time, or will you be left with a cheap dead snail.