https://github.com/ClassicLootManager/ClassicLootManager/blob/c1427c50a0b733a86aacadf7e81bcaecc6796d82/Modules/RosterManager/Roster.lua#L255+5

Could the section of code use the minimum GP inside the calculation of the decay?

Expected

Player with GP score of 300 and with a roster configuration setting Minimum GP to 250.
Applying a 50% decay will result in the player having a new GP of: 275

Actual

Applying a 50% decay will result in the player having a new GP of: 250

Outcome

As implemented, we had a situation where player1 had a higher PR than player2 before decay, but due to player2 having higher EP, and both players bottoming out at the set minimum 250 GP, the PR order reversed.

CEPGP provides an option to decide how decay is applied with Minimum.

Well you might want to do the math but that is a bit of oversimplification.
Basically if you don't want to loose PR i think the gp is calculated something like this
GP_Spent <= 1/PR * (EP_Gained / (1 - Decay)), where if they are equal then your PR for next week is kept intact.
But that's just from my quick napkin math, can't guarantee no errors there.

This topic was heavily disussed and calculated when we decided to not introduce any misconceptions introduced by other addons. https://docs.google.com/spreadsheets/d/1IhgEbGQifwWprD5ejJ5mYMyPmq7Ue3F16xDs1gP1z4w/edit?usp=sharing As you can clearly see (GP1 approach is what you are saying, GP2 is what we decided on) GP1 based prio behaves in unexpected way.
Proper decay math, as designed in EPGP system nearly 15 years ago is GP[n] = max(minGP, (1-decay)*GP[n-1]). This also follows common sense of decaying the actual value and you see how you decay it (until you reach the threshold).

In addition your described case is great example of how decay should work:
For the PR swap to happen there is a specific point setup needed: Player with bigger PR should have more EP and less GP than player with less PR. This mean that the former player is hoarding (getting EP, not getting GP). And the decay hit him more as a natural protection to hoarding. I see a win-win situation here.

Thanks for your reply, and help in understanding the implementation decision.

It was an oddity to see though, and trips up our raiders where we're all only waiting for BiS items atm.

If I may re-frame your response so it's communicable to my raiders;

The minimum GP * decay% determines the amount of GP one could spend in a raid week without it affecting next week's PR.

Would that be appropriate?

That spreadsheet is quite incorrect as far as describing how Cepgp decay functions with decay-free threshold enabled. This is how it works on GP1:

https://docs.google.com/spreadsheets/d/1NcR4s2RNdTNU8mWJC9NgDvHZd6QqZltT-l3ovP2DWFw/edit?usp=drivesdk