Measure legislative independence, frequency of regime changes, election margins, concentrations of executive authority and so on. Sure, it's a 'political opinion' that nobody is so cool they are naturally re-elected over and over with 97%+ majorities, but you can certainly measure the number of standard deviations in election results.
This book is readable and essays a rigorous approach to the topic, albeit within an existing political science/international relations framework whose axioms are not universally agreed upon.
Most of these things cannot be "measured", and weighing them to produce some kind of aggregate score is inherently a biased process. There isn't any remotely agreed upon method to determine what constitutes an authoritarian government (or even where a specific government falls on a scale between authoritarian and democracy, when compared to others).
In some aspects, Switzerland is more authoritarian than Saudi Arabia. For example, in Switzerland there are strict building codes everywhere that restrict what kind of house you are allowed to build for yourself; in most of Saudi Arabia, you can build your house however you want. The very idea of a scale that somehow "measures" such things, and adequately incorporates them into a coherent picture of the whole, is absurd.
Any evaluation process is subject to accusations of bias, including yours above:
Instead of asking whether or not XYZ is a dictatorship, ask "are they following their own laws and constitution?", "are they respecting universal human rights?", and "in whose interests are they acting?"
Instead of just nay-saying and trying to redirect the argument, you could try engaging with the question of what a dictatorship is as a political structure. Or not, as you prefer.
This book is readable and essays a rigorous approach to the topic, albeit within an existing political science/international relations framework whose axioms are not universally agreed upon.
https://mitpress.mit.edu/9780262524407/the-logic-of-politica...