It means there are no real "concessions" in a sale... but the concession is a nothing burger in the SUBJECT - only an issue in a comp. OP seems to be having a problem differentiating between the Comp and Subject concessions.
Just by pulling it out your .....?
If you cannot support the adjustment with a paired sale or perhaps sensitivity, then state "insufficient market evidence to support an adjustment" and provide a general observation in the reconciliation. "I felt Comp X had additional amenities over Comp Y therefore, it was more similar to the subject and I gave additional weight to Comp X. "
Making small adjustments is needless and can be misleading. If you are making more adjustments than necessary it is pointless and if you are making 10-20 adjustments in the grid, then your numbers statistically are simply a random number generator being generated non-randomly by your skew.
I've yet in 30 years to make an adjustment for porch, deck, style, etc. I found a better comp, or I simply ignored it. So, when I started doing multi-linear regression, I discovered that even a fireplace is nigh impossible to isolate numerically. I rarely adjust for more than four differences. Land -value, first and foremost; outbuildings, pools, etc.; and finally size and effective age (condition-age). If that doesn't tighten up the grid, then I need better comps.