ELI5: Reservoir Volume Estimation Probability Questions

I understand that P10, P50 and P90 values are derived from Monte Carlo simulations. A P50 volume is the size at which 50% of results are above and 50% are below and the most likely outcome. We've used these 3 P Values to try and create a lognormal distribution to be able to approximate where for example the P63 volume is or probability of where there is a 63% chance of there being "at least" that volume of resource (or some useable estimation if they are separate things.) The problem I have is that I personally don't quite understand the solution we have and want to hear from a petroleum or reservoir engineer about what the correct terminology is and the best way to go about things. My main problem is that I see graphs around with cumulative probability distribution with the P10-P90 values taken out at the relevant 10-90% y axis. But using P values to create a lognormal distribution kind of means that the 10-90% levels on the graph aren't P Values any more but cumulative probability... (at least as far as it's been explained to me which may or may not be correct). So when I put in a P10, P50 & P90 even though the P50 entered to create the distribution might be 40mmbbl the 50% cumulative probability comes back 30mmbbl depending on the standard deviation of the other P Values or how skewed the distribution is. It's obvious that the P90 isn't a 90% chance of that figure because it's on the far side of the bellcurve but I HAD thought that P90 meant "simulation says 90% of eventualities were above this figure and therefore there is a 90% chance it's AT LEAST this amount." so not that amount, but that amount *or more*. That's what I've thought for a long time. But using the 3 P Values on a distribution creates very differently weighted curves for each set of P Values which pushes the results out unless they happens to be on a (roughly)1:2:6 ratio like P90: 100mmbbl, P50: 200mmbbl & P10: 600mmbbl.. Then the values from the curve at the 10, 50 and 90% points come back very close to the 3 P Values entered. Basically I sort of want an "Explain Like I'm 5 years Old" of why the P50.. The point at which half the results are above and half the results are below, and supposedly the most likely scenario.. isn't the same as the 50% cumulative probability on the curve.. A large part of my problem is that I think in visuals and I haven't seen a graph of the results. I sort of understand that when one or two of the P Values are a different standard deviation away than the others that it skews the curve along the X axis. I'm just not intuiting why P50 isn't the same value as "50% chance of at least this much" on the curve. Either that or we're not doing it right. Or that when the 10, 50, 90% values are different to the P Values used to create the distribution, the difference is a measurement of deviation.. or something.. I don't know. All I want to do is use the P10, P50 and P90 mmbbl values to get the statement "There is 43% chance of at least Xmmbbl" and be confident that it is correct and a petroleum engineer won't look at it and go "that's not how it works" and that I can explain it to someone else.. proving that I understand it. So I guess I just need an explanation to clear things up, and maybe to come back with a couple more questions until I get it and how it's supposed to be done..