VinDSL -- aka VD -- is gonna love this, but from the posts on this thread, I feel comfortable with this assessment.
Companies go through ascendancies and declines. It can be associated with changes in personnel, or rapid hiring of personnel.
It may be that the Ultima 90 was designed as a little brother to the Ultra-120-X, just as the XP-90 and -90C was a remedy for XP-120 issues with some motherboards.
I see AigoMorla and some others have no problem mounting the sink, and this may be motherboard-specific.
The fellow at TR who provided the "tech-rep" response as quoted in an earlier post on this thread had some "King's English" issues, and seemed to lack experience with our problems here or how to deal with them.
On the matter of lapping and TR warranty, the TR reps to whom I spoke were never explicit about "why" lapping would void the warranty, except to say that the nickel-plating of the base and pipes aided in preserving the integrity of the solder-joints with the pipes.
They chose Nickel for the plating material because it is right next to Copper on the Galvanic Table of metals. Metals on the Galvanic Table that are widely separated from each other in their electrical conductivity features (at least, that's what I recall from reading up on Galvanic corrosion) -- tend to promote corrosion (on one metal or another) if put in contact with each other in an electrolytic solution. Metals that are close together on the galvanic table tend to have minimal corrosion under those circumstances.
Taken in conjunction with the remarks about the solder joints as made by TR personnel, this apparently drives their warranty policy.
I discussed this with SVC personnel prior to ordering a "custom-lapped" TR-Ultra-120-Xtreme, and we seemed to have immediate agreement in our speculations and conclusions. TR is afraid that removing the Nickel from even the bottom of the heatsink base will cause the rest of the Nickel to eventually peel away and compromise the heatpipe solder-joints.
If one were to worry about this, you should only need to put a thin layer of thermal epoxy on lapped copper surfaces that are exposed to the air and do not come into contact with the CPU heatspreader. But I cannot believe that the corrosion of copper in exposure to the air is bound to be that quick, or even likely, unless you have a house that is right on the beach, or run your computer in your yacht -- where the air has a high salt content.
The SVC tech-rep seemed to agree, and eagerly noted that SVC would warranty the product on the custom-lapped version that they ship.
But for the convex-cylindrical heatsink base, the issues of the mounting mechanisms going through revision after their product hits the market, one must conclude (a) the convex base design is a design-flaw bordering on silly, (b) for their innovative design record for performance only, they are in too much of a rush to get their products on the market and are not devoting enough attention to testing.
So -- even I, BonzaiDuck (you thirry thavageth ) -- can say "BUyer Beware," but from a different perspective: If you consider buying, become familiar with these drawbacks first. You may want to consider whether you want to go to the trouble of "modding and lapping" and whether it is worth your time and trouble. It depends on how much leisure time you have, and how much leisure time you want to spare. If you run into trouble with a particular motherboard, you may have to make corrective mods to the mounting mechanism, which means more time and trouble.
This is all really a shame, though, because the U-120-X and even the Ultima 90 to a lesser extent have proven performance potential. The design flaws would drive many backward to second-best -- nevertheless "close-runner-up" -- choices. But why the hell do you have to put up with a need to do DIY mods on the heatsink to get it to perform at an optimum?
Ridiculous.
Per the other poster who was "worried about his Core temperatures." Those load temperatures seem about par for the course -- in the 60's Centigrade across the cores. I can imagine a few degrees C difference in performance between the Ultima and the U-120-X, on the basis of size -- favoring the U-120-X.
My own Q6600 (B3) implementation for the U-120-X shows temps with the hottest core around 54C at room-ambient 75F in a 33% OC (1:1) to 3.0Ghz. At 3.2Ghz, my last testing showed peak values of around 64C-65C at 79F room ambient. The 3 Ghz OC has VCORE at 1.3V, while the 3.2 OC requires VCORE = 1.4188V. Keep in mind, though, I'm using diamond TIM (worth an improvement of a couple-degrees), and a motherboard-ducting mod. The ducting mod probably goes farther to cool mosfets and chipsets, because the TR-U-120-X is highly efficient and so ducting it may only be worth 4C to 5C improvement in CPU temperature alone.