OK how good is the Unurium production of the new wonder? Let's do some rough math with the numbers presented here.
My aim is to compare (a) building the new wonder to produce Unurium with (b) building festival merchants and producing seeds instead; and balancing that by switching production in the trader.
Chapter 19 Magic Residences produce 1.1M$ per day. Let's say I have 25 of them with 150% culture level, that's 1.1M$ * 25 * 1.5 = 41M$/day.
A level 1 [or 6 or 30] Vallorian tower will produce 0.024% [or 0.048% or 0.16%] of that money in Unurium, that makes 10k [or 20k or 66k] per day.
My trader (with 5 collections/day) can produce around 200k Unurium/day in chapter 19. If I want to produce the 10k [or 20k or 66k] in the trader instead of the wonder, that eats up 5% [or 10% or 33%] of its daily production time.
I would take that production time away from seeds. My trader can produce around 100k seed/day, so that will cost me 5k [or 10k or 33k] seeds per day.
With festival merchants producing 3840/day, this means I need 1.3 [or 2.6 or 8.6] merchants to make up for the loss of seeds. At a size of 4x3, they will need 16 [or 31 or 103] squares of real estate. The Vallorian tower has 5x6=30 squares, so its production per square starts worse than festival merchants and overtakes them at level 6.
So much for pure Unurium production.
Switching to tournament perspective, each wonder level raises spire and tourney squad sizes by 0.3%. Already having 360 wonder levels means that a level 30 wonder raises my wonder-based squad size factor from (1+ 360*0.3 = 208%) to (1+ 390*0.3 = 217%) for an effective rise of troop losses by 217/208 -1 = 4.3%
Compare that to the extra merchants needing 103-30 = 73 squares = 3 expansions more, which you could save (and simply not build) if you raise the wonder to level 30 instead.
Already sitting on 122 normal plus 39 premium expansions means that 3 more normal expansions raise my land-based squad size factor from (122 + 0.66*39 - 4.5 = 143.24) to 146.24 for an effective rise of troop losses by 146.24/143.24 -1 = 2,1%.
2.1% versus 4.3% means that... using this wonder is TWICE as damaging to my tournament losses compared to the alternative of buying expansions and building merchants on them.