Thanks for showing us the heatsink without the logo. With the logo:
https://www.gigabyte.com/us/Motherboard/GA-Z97X-Gaming-7-rev-10#ov
Hard to tell what's going on there. Regardless, I was thinking something closer to:
https://www.asus.com/Motherboards/P5Q3/gallery/
Especially that hefty heatsink there in the middle. With even a little case airflow, that thing should fit the bill. TDP for P45 chipset was 22W. Granted, that was probably split between Northbridge and Southbridge but still. You get the idea.
This is a p5q-e heatsink shared between the left set of vrms and the chipset. The p45 chipset per my caliper measures 11.5x9.5mm.
The heatsink seems stronger that it really is. The deceiving color makes you think we are talking about 100% copper but in reality is anodized aluminum made to look like copper. To make things worse, the joint between the fin stack and the baseplate making contact with the thermal pad is not soldered but glued (i found out by breaking the top vrm heatsink of said motherboard
These kind of heatsinks banked big on you running stock cooling configurations, which for intel lga 775 were the aluminum extruded radial design with or without a copper core. This is why so many fins make sense as this heatsink is integrally dependant of the air renovation made by the open side nature of the intel fan.
Now, moving to 2019, where overclocking is mostly a feat made by big tower coolers and AIOs which criminally neglect chipset and vrm cooling, with a chipset that is purposedly far away the cpu socket so it stays near the pci-e i/o for better signal integrity required for pci-4 routing, and thus need low profile solutions so it doesnt interfere with the add in pci-e cards you would use on those ports
You see that this sum of factors make the natural conclusion be different from the p45 era and now? Even with a fin + heatpipes shared heatsink the design would fail because more than any other design, fin stacks need forced ventilation to make it work, as the total volume of metal is too low to dissipate the heat. Forced ventilation on enthusiast boards tend to be neglected by the nature of the cpu cooling involved, even the wraith spire with its closed side design and ~90mm height isnt that close for it to provide correct airflow to an hipothetical retro vrm+chipset design.
Another thing to consider is that whereas resistance of goold old 200x year mosfet was really high compared to the low rds "meta" of today, the surface area of said ics was bigger. Now we are packing integrated mosfets in such small places that fin stack designs might prove difficult, or worse, provide worse temperatures as fin stacks depend on baseplate surface area to transfer properly the heat to the heatsink. My ab350 gaming itx for example has only frigging 6.5mm of width to place a vrm heatsink, the most common heatpipe width is 6mm diameter, how can you achieve a heatpipe + fin stack design on such confined spaces.
I know pcb design could be different to accommodate better retro vrm+chipset heatsinks, we even have from time to time somewhat lookalikes like the z390 fatality itx which even if it doesnt use fin stacks, it shares a heatpioe between vrm and chipset so the vrms bank on the chipset heatsink surface area to better dissipate heat.
In the end economics of the scale favouring extruded alu heatsinks also play a factor. But these economics became possible because the industry altogether moved away from tini fin stack vrm heatsinks.