I've glimpsed over
your cited paper and their methodology to try to identify individual miners who point their hashpower to pools looks quite legit (to me). It's not fool-proof but to my understanding they tried to apply reasonable efford to make a good attribution.
It's not too surprising that in Bitcoin mining bigger is better as you gain advantage with size because certain cost factors don't scale linearly with size. A very simple example is an individual miner with 10 ASICs. This individual could likely handle also 20 ASICSs alone, so no additional expenses for labour. Maybe this miner also has the space for the additional 10 ASICs, so almost no additional cost for storage of his mining rig. On the other hand he has doubled costs for energy and likely cooling.
I'm very much simplifying here…This confirms that mining is very centralised and controlled by a few institutional investors acting in their own interests and not those of the other participants. Is there a chance to know who are the investors of Foundry, AntPool, F2Pool and ViaBTC?
How do you come to that conclusion regarding your institutional investors?
Hashpower concentrates into pools. According to your cited paper there are not many miners with a great percentage of hashpower that point their hashpower to large mining pools. Agreed on that and this development isn't actually too surprising and I personally don't see much of an issue here. It's economy in a very competetive playground.
In theory everyone can join the network, but in practice those who actually participate in the PoW consensus (51%) and produce blocks are less than 0.1% of the miners, i.e. those 50-60 miners reported in the research.
This is nonsense. In practice anybody can join a mining pool and point his hashpower to it. You're probably excluded from certain larger pools that have a lower limit of required minimum hashpower, but that doesn't prevent you from finding some other pool that will accept your hashpower.
Are you sure, you understand PoW mining? There's no such thing as PoW consensus (51%). Bitcoin mining PoW forces you do execute hashwork to find a blockheader hash that satisfies the required mining difficulty (your to be found blockheader hash needs to be lower than a certain hash value dictated by current difficulty). Finding such a blockheader hash is due to the used hash algorithm a completely random process. You can hit a valid blockheader hash within a few thousands or millions hashes (statistically very … very unlikely) or you need a ridiculously high number of hashes to hit it for which you need longer than some other miner who was luckier than you.
Again: it's a random process and with higher hashpower you gain statistical advantage.
Mining pools mitigate the risk of solo-mining. With "fair" pools you're paid statistically according to your provided hashpower within the granularity of the pool's payout scheme. Choose your mining pool wisely!