Even with their (many!) limitations I wouldn't just discard Chinese oscilloscopes. Honestly, it makes me feel bad to have a Rigol and read those fantastic application notes from Keysight, Tektronix, etc.
An example of how useful (or not) the infamous DS1000Z can be.
Last week a friend was struggling with a simple communications problem: reading information from a device using Modbus. When he asked for help all he could say was "the reading function returns a timeout".
How do you debug that? Of course there are two manufacturers involved and they will probably say "my stuff is fine". So you grab the cheap DS1000Z, make a capture, and you find out all this (warning, not rocket science, pretty obvious!)
- The command is properly sent and understood by the other device. It actually replies. Also, it replied in 25 ms.
- There is a bug in the Modbus library of the device sending the request, it doesn't set the RS485 interface in receive mode after transmitting.
- We better add termination, although the signal is useful better to avoid that ugly overshot (the decoded screenshots were taken after we fitted a couple of resistors)
- And we could even decode a couple of packets in order to check that it really works according to the documentation.
With all the deficiencies, "pluses" and a somewhat provincial software design with subpar interface responsiveness, and surely it helps that this is not exactly rocket science (Modbus at 9600 bps over a 50 m twisted pair), I would say that the 450 euros of the oscilloscope have been more than worthy.
While we would have found out sooner or later, the toy really allowed us to find out pretty soon. Moreover, you can always make a better case with the manufacturer when you can document your findings.
I know there are much better oscilloscopes which are much more expensive. I now that the math functions in my old LeCroy 9400 are much better designed. But what was the price of a 9400 back when it was launched?