Thanks Giuseppe for bring this up: I had overlooked it on its first submission.
He talked about market ecosystems of tightly integrated companies coordinated through the use of technological applications (with AI among the others) that optimize processes and market adoptions with the use of big data set.
(I cite him by memory... and I hope I didn't misunderstood what he meant)
In my personal view (apparently pretty similar to your and Roberto's one in the effects), companies are more like unicellular organisms that operate in a multidimensional environment with money as the form of energy, people as a sort of DNA and their output and exchanges as other aspects of their biological lives.
So to my eyes, Google becoming Alphabet (which includes and coordinate several companies) is a sort of evolutive step towards what in biology is a multicellular organism.
Now I know nothing about biology and this model might be just crazy stupid.
Still if any of you have some research to share about lives of unicellular organisms (eg bacteria) please send me a reference since I've been unlucky with my google-fu.
My insight is that we could use biologists knowledge to explain companies and multinational behaviour on a world scale.
I find this model useful to understand the relationship between companies and people (employees, customers, users, politicians and so on...), companies and states, companies and environmental issues and so on...
Particularly interesting are the relationship between companies, where you can clearly see predations, symbioses and even parassitisms.
Recently some companies grew bigger than any others, even more powerful than many states (actually Hobbes would probably define them Leviathans on their own), and they turned dangerous for the whole environment, more or less like a tumour.
The process is still ongoing and no equilibrium emerged yet, but whatever the model, most people are not perceiving the risks.
Given all that, I agree that the data taxation is not resolutive of the problem.
But are we sure we understand the problem?
Personal data taxation, if designed correctly (based on each bit use), might slow down the process enough for people to find better models: the purpose of economy (and politics, and...) must be to serve humans, not to feed companies.
On the other hand these companies were born and grew in a paricular technological environment. And a rather clumpsy one, around the abuse of HTTP and browsers.
Google knows well how suboptimal and unstable is the stack they have colonized, that they try to lock people as much as possible, fostering a browser growth so much that they can literally run hardware emulators and operating system over them!
So I agree with Philippe too: the easiest way to get rid of the power of these companies is to disrupt their stack with better technologies.
The fediverse in my opinion is interesting but still built upon the same stack.
Unfortunately few are brave enough to try something really new (and even fewer in Europe).
Why? Fear of failure, but also no support from the environment (which lacks vision or competence).
Giacomo
PS sorry admins for the long mind dump...