I'm looking for options before the build. 

I use whatever I have on hand. Usually Thermal Grizzly Kryonaut but I'll use any good brand thermal paste like Noctua NT H2, Cooler Master MasterGel, Arctic Silver 5. It doesn't really matter all that much to be honest. You're usually talking margin of error difference between good thermal pasts in the range of one degree one way or the other. 

Using Artic MX4 here.  Temps are pretty decent.

I usually use Prolimatec PK-3
 

Here's a more recent table from Tom's 2021 testing (they seem to have a fairly large list):

(might have to open image in new tab)
Comes from this article.

Their older test (2017) did show the diamond 24 doing quite well:

It doesn't look like they tested the Diamond 24 on their newer test.
 
I do find it interesting how little difference most of these sites show and one thing I think is generally missing is technique. The different viscosity pastes lend themselves to different application techniques. GN did a good video showing some of the differences using one type of paste, but they didn't really get into how each method works or applies the advantages of one method over the other when you look at different viscosity pastes. For example, the Prolimatech PK-3 and the KPx are very different viscosity. I know a lot of people go for the spatula full coverage spread (before putting the cold block on), but I have found that the KPx does really well on a 12900k or 13900k with a small line down the center and 1 very small dot in each of the 4 corners. It spreads really thin but still covers everything and does well. I think the PK-3 could respond really well to pre-spreading it very thinly as it's super thick stuff (much drier feeling than KPx). That said, I just use the same method on these. I'm not sure if I just haven't tried the pre-spreading method enough to work out the kinks, but it always results in way too much paste and a mess for me. All that said, I think most of the reviewers use the pre-spread method for their tests as GN does and they point out that they somehow do better with consistency that way. I always get better results with a different method, whether that's the thin line (used to be the grain of rice on square CPUs) or a small oblong blob in the center (small-pea-sized). 

Digital_Fuzion
Great information! What is your all's thought on using a contact frame to lower temps even more? 

 Intel 12th Gen contact frame for socket LGA1700

Honestly, while it SEEMS a bit "gimmick", testing has shown that contact frames (in general) generally result in a more uniform thermal interface with your heatsink.

Now, is it "enough" to make you spend $44 on a full-fat Thermal Grizzly unit?
Or will a $17-25 clone be sufficient?

That's a value judgement on your part.

TG is doing pretty much ALL the research, and the clones are just duplicating specs to a certain degree.
PERSONALLY, if someone's doing all the work to make your experience "better", I'm inclined to give them my money.  Even if it means I'm paying more.