I love hydraulics; no surprise there. If there is any application where I feel hydraulics is at least a good fit for a machine, I’ll be elated to make it work. However, hydraulics is not always the best choice for a drive and control application.
If there is an application more easily driven mechanically, then I often concede so, especially if efficiency is a consideration. A couple examples which come to mind (and these are actual requests) are a bicycle and windmill.
The inventor of the hydraulic bicycle pictured a small hydraulic pump connected with hoses to a small hydraulic motor. It’s a nifty idea, to be sure, but would ultimately result in epic fail. The pump would be driven by the pedals and attached via hoses to the motor. In practice, it would require a small reservoir to manage heat and air.
Besides adding mass and complexity to replace an already compact and light gear and chain arrangement, a hydraulic bike drive would decrease the output on the drive tire relative to your pedaling. An 80% efficient pump turning an 80% efficient motor would give you a 64% efficient package. If you pedaled at 100 RPM, you’d see 64 RPM on the bike wheel, with even less torque than would be applied via a chain.
Another individual wanted me to design a hydraulically transmitted windmill. The prop would turn a hydraulic pump and power would be transmitted down a hundred feet of hydraulic hose to a motor powering a generator. The same inefficiencies present in the bicycle would affect power output at the generator, but with the added backpressure of long hoses and the heat generation cumulating from a high duty-cycle application.
As you can see, hydraulics is not the answer to every power transmission application, especially when a straight mechanical option is provided.