If you got any improvement at all, you’d be doing better than expected. The Official Party Line is you can’t “add acting” and you can’t change one voice into another without odd results.
If you got something you like, then you win.
You should do one other thing. If you started out with MP3 (a common mistake) you should produce your final show file before you call it good. Audacity doesn’t like making MP3s from other MP3s and you could get honky compression problems.
Acting is indeed the best way to create a commanding voice. However, there are some things that can improve the effect by post-processing.
Compression is not a thing I would use for this, quite the contrary.
Accents ar very prominent, so use an expander.
The overall volume goes down, therefore we need a limiter to even out the higher peaks and bring it all up again.
Consonants are very accentuated. Either use the eq for boosting frequencies over 4000 Hz. or duplicate the track and apply a highpass filter.
The speed is comparably fast (I’ve used 77 % for the example below)
The overall length of the speech stays practically the same, therefore, the pauses must be longer to produce a slightly choppy effect.
The end of sentences do scarcely go down in pitch (below, the “once” is raised by a half tone).
The overall pitch is increased (at least if I think of my women in the family…)
Remember those boot camp sergeants? They stand right in front of you and shout like hell. This means that the direct sound is much faster than any reflections from the wall, thus a pre-delay of 50 ms isn’t unusual.
In general, anything that creates presence is welcomed (e.g EQ boosts for 315/3150 Hz and a cut at 1000 Hz).
Here’s the example that illustrates some of these effects a bit.