Biopsies will, of course, report the highest grade score. But, considering a needle in your walnut sized gland, even up to 20 times, still only represents a fraction of the tissue available then you are still dealing with a very small sample. The post-surgical pathology is far more accurate but shouldn't be lower than the cores from your biopsy.
But, you mention something that I also mentioned: "unless the pathology on the biopsy was inaccurate - which can happen". The pathologist is rarely the surgeon, so the surgeon might grade you differently but the pathologist is likely to be more accurate. But, even if all things are equal, it's completely possible for two qualified persons (of whatever discipline) could grade the samples differently - it happens from time to time.
So, take your examples. You had 70% 4 + 3 and 10% 4 + 4 - let's say it's off and it's 70% 3 + 4 and 10% 4 + 3. While the severity of the cancer is less, your need for treatment isn't - either way there is a lot of unfavorable cancer in your prostate. I had just 5% 3 + 4, and would have been on active surveillance if not for the 0.68 Decipher. Since my post surgical path was >30% 3 + 4 it was the right thing for me to do.
Now if you were 3 + 3 and 3 + 4 and that was reduced to simply 3 + 3 then that's a whole different ballgame since 6 is not even considered cancer by many urologists.
I think the point I am making is that we all want to grasp at whatever bit of good news we can, and questioning the pathology of the biopsy is an easy one because it's known that two people can have different opinions, but in some cases it doesn't make a lot of difference. In any case, you have to be your own advocate and if you believe the results to be inaccurate then seek more consults and new pathologies and be absolutely positive. There's a flip side to all of this, if we want to believe the second pathologist's lower rating and then use that data to not seek treatment and that person is wrong then it's game over. In the end, we choose to determine which person is correct, right or wrong.
I am of course not bsnking on a downgrade, but it is an additional source of diceyness in the decision.
And there are several studies that look at doengrades and upgrades post-pathology. And with 4+4 they find 40-50% probability of a downgrade and something like 10% of an upgrade. On the other hand, lower grade cancers at biopsy have a higher chance of being upgraded.
Two things are at play here, I believe. Firstly, regression to mean: Any estimate that is to either extreme is likely to be an over or underestimate. Secondly, you use a different metric. The max of a small, untepresentative sample at biopsy and the median at pathology.