Inductor Accuracy

[vk4djf]

I'm rebuilding my classic band RF board. I've got the enhanced schematics with the listings that include the exact inductor values. I've started with the 80m and 40m inductors. I noted I could get the 80m ones to within 5%. I tested them with the inductor isolated just sitting on my bench, then again when the soldered in. The inductor soldered in is 5-10% more inductive when soldered in.
Is it a valid test to actually measure them soldered in place? I'm assuming the target inductance is when in place, but I'm not sure I can measure that correctly.
Secondly something I discovered was variance between cores. The 80m one that was 676 I got to 714 at best. The 40m one that was also 676 with the same core type and same number of turns came out to 569 at best. Couldn't get it any higher than that.
Is it normal to see that much variance between two supposedly identical cores?
I'm feeling like I need to rewind some with less or more turns to get closer to what they're supposed to be.
Anyone else done it this way to get better efficiency out of your Trusdx build?

[LILO]

Waste of time. Just follow the well-known tuning procedure. That's all you need.

[vk4djf]

It's not that well known. It's in a slightly vague video and just suggests minor tweaks of alignment of the turns. Last time I did this I wasn't able to get anywhere greater than 4W and poor efficiency on any band.

[LILO]

How do you measure them on which frequency?