re word your question and add some more information, that probably the reason why no one has replied

OP here. I'm not able to edit the original post, but I'll try to explain what I mean. Here are the steps I have done:
1. I calculated the allowable pile capacity using Excel, using the alpha method (weak clay). I got:
Allowable Capacity = 548 KN
2. Since the piles are 2.5 diameter close to each other, I calculated the Group Efficiency using Los Angeles method. I got Reduced Allowable Capacity = 453.59 KN.
3. I estimated the pile to be a point spring using the silly equation everyone is using ( K = EA/L, E from concrete). I'm not enthusiastic to treat the pile stiffness as coming only from concrete since I have no rock at the bottom, but I have no time currently to investigate more into better equation. I modeled the structure in CSI Safe using the point springs.
4. I got the max pile reaction to be 460.89 KN, exceeding the Reduced Allowable Capacity (Reduced due to Group effect) but less than single pile allowable capacity. My question is: Are the piles OK? My engineering feeling is mixed: not all piles are that much stressed in a local neighberhood. Say one pile is stressed beyond the group-max-allowed capacity but the 4 piles around it are stressed for less.

To reword more my question, shall I judge a single pile for exceeding the group efficiency or should I average down local groups of 4 (2 x 2) for the comparison?

Ok, thanks for the further explanation. To me 453.59 and 460.89 are essentially the same, given the level of accuracy we work to. So I woulndt give it a second thought.

But if you are submitting this to a pain in the ass reviewer then they might reject it and say get it under 453kN. (I often get similar issues with reviewers saying 25.2mm of settlement is not acceptable, has to be 24.9mm max.....)

One way to get around this might be to state that the lowest pile capacity in the group is 460 which is a factor of safety of 2.89 (or what ever), this is considered acceptable as load testing will be undertaken to verify the capacity of a single pile.

I would hope that you are intending to do this.

But to be honest your biggest unknown or error is not using an appropriate pile spring. Using PL/AE is elastic deflection of the concrete only, as you said. No allowance for soil settlement. So you are over egging your pile stresses. If your piles are too stiff they will attract more load right. Not sure how your loading is in your model, whether its uniform or non uniform but this will have a big effect.

Thanks EireChch. That was helpful. I have submitted the data already, but it's very important to get things right for the next time.