Nice that you try to point with some number examples but your example is pretty far off from the reality if you compare them with statistics from other (similar games).
Let me take a second try at this.
I´ve chosen 3 games for comparison:
The Elder scrolls online [A classical AAA MMORPG]
EvE Online [Massive "One Server" Sandbox environment MMO]
Rust [Similarities since both are about building structures in a finite world, no connected servers though]
Checking steamspy.com for the number of activated copies on steam (i know that this source is not the most reliable one,but certainly better than guessed numbers )
Activated copies on steam
ESO: 503,980 ± 17,109
EvE: 383,701 ± 14,931
Rust: 3,734,572 ± 46,324
Peak number of concurrent players for each game from steamcharts.com:
Peak player number
Checking the websites of various games that share the number of their total subscribers/ actively playing accounts (DotA, EvE Online, and some others) and comparing those numbers with the peak number of active players I approximated that the number of active players roughly translates to 10 times the peak number of players (Which goes along with the number of active players displayed on steamspy.com (2 weeks observation period))
Which results in this approximation:
Approximated total active players
I couldn´t believe the number on Rust myself but several websites confirmed it.
Percentage of active players in comparison to the total number of accounts
(Bracketed values are from steamspy.com)
ESO: 6,9% (7,8%)
EvE: 9,5% (12,2%)
Rust: 13,4% (13,5%)
Or in pie charts:
And those numbers just show the total number of active players (including old & new ones) so the percentage of, what you call, "Permanent players" would be even lower than that.
Considering the similar values for this (very) different games I´d assume that this will also come very close to the numbers in Boundless.
Translating into beacon numbers
This discussion is about inactive beacons, not inactive players after all so I tried my best to translate my results into a number of beacons.
I combined your assumed percentages with my 'activity results' and the result of the Beacon survey result to hopefully get the most neutral and fitting result.
(I just summed up your "sprint players" and "displeased players" and extrapolated the percentages to a whole of 100% for easier math and used very basic excel and linear interpolation tools for the results.)
(Based on the results from Rust to get the most 'optimistic' results)
(Assuming that the average "active player" has played around 100 hours (≙ 111 beacons))
So after the initial hype around 50% of beacons would (could) be owned by inactive players.
I don´t dare to conclude if this would justify not-permanent beacons (although personally I think so) but it would definitely be mildly annoying if every 2nd shop in a city would be abandoned.
PS: These are the current stats of Boundless according to steamspy.com: