Well, there's definitely a bug in the loot frequency formula
https://plotly.com/~therainycat/53.embed
I'll check it later, but from what I can see, a reasonable move would be to reduce the loot drop rate at Frequency above 1000% by around 30% and smooth out the curve to increase loot drop rate for Frequency below 1000%.

Once again, nice find. I rarely play at such a late stage of the game (usually because I run my saved games) and it is hard to spot such a sudden change in a loot drop rate.

I've collected some extra data.
These are 4 slices with a different average difficulty of portals, endless mode:
https://plotly.com/~therainycat/63/#/

There's a bug in a drop rate calculation formula which is directly related to the average difficulty of portals - just like in your table, the rate starts to grow significantly only at the highest values of LOOT_FREQUENCY research and that's not intended.
For a reference, there's also a line which shows the intended drop rate (1:1 to loot frequency, precisely as stated in the research) and it shows that it "works as expected" only at values of 100% loot frequency and around 1050% loot frequency. Loot drop rate is significantly lower than it should be in this range (600% LDR actually gives 300% of loot) and I'm going to increase it, but it also means that loot drop rate at values above 1050% will be significantly reduced.

By the way, these charts show the amount of loot compared to 100% loot drop rate. They are being multiplied once again with the average difficulty of the map, meaning that difficulty affects the drop rate twice (exponentially). I'll leave the part which multiplies loot drop rate depending on an average difficulty, meaning that higher map difficulty will still increase the drop rate, but LOOT_FREQUENCY itself will become linear and will work exactly as the research states.

In other words - loot drop rate will be increased by up to 2x with LDR values below 1050% but it will also be reduced at LDR values above 1050% (1120% is a maximum achievable value) by up to 4.5x. It will become easier to farm loot and players will get faster to 1050% LDR research (prestige research level around 92/100) but loot will drop up to around 4.5x less frequently on a maxed out endless research tree (LDR research levels around 93-100).

Edit: replaced tildas with "around" because Mantis.

Forgot to mention:
I've also compared the areas of each chart - considering a linear progression through the loot frequency research (for simplicity), the area should represent the average effect of this research. In practice, players will probably spend more time (per upgrade level) progressing to level 92 of the Loot frequency research so the resulting buff may be even more significant than stated below.

Area of the charts:

• Baseline (linear growth of loot frequency / the chart I'm going for in 1.9.0): 141220 (100%)
• 1500 AVG difficulty: 125521 (baseline x0.88)
• 500 AVG difficulty: 115510 (baseline x0.82)
• 200 AVG difficulty: 110400 (baseline x0.78)
• 50 AVG difficulty: 106757 (baseline x0.76)

Since formulas will be adjusted to be closer to the baseline, this will increase the average loot drop rate by up to 31% (1/0.76) at all loot drop rate research levels.

I've decided to keep the loot as it is. Too much effort, time and risk for a thing which has been there for years already.
Not touching the economics in this game.