Most published monsters seem to involve some amount of 'game feel' balancing based on designer's intuition or playtesting rather than following the calculations directly. These 'game feel' assessments should occur when crossing between CR's 4 and 5, CR's 10 and 11, and CR's 16 and 17, corresponding to the boundaries of the player tiers, due to conditions, spells, or instant death effects that may be impossible to deal with at lower levels or trivially easy at higher levels.

Then, barring any independent balancing steps you as the DM might deem necessary due to 'game feel', the math should work out correctly. The math is reversed when sliding to a row of a lower CR. Since each of the n-many rows increased the CR by exactly +1, you've increased the monster's CR by + n.This new average CR is the estimated CR of the adjusted monster. Thus, you increase the average CR by n-many rows in the table, since the average has increased by ( n + n) ÷ 2 = n-many rows.You increase the defensive CR by n-many rows in the table.You increase the offensive CR by n-many rows in the table.When you slide the creature's statistics by n-many rows to a row of a higher CR in the Monster Statistics by Challenge Rating table in the manner you've described (including all of the statistics covered by the table, not just some of them), the practical effect on the CR calculation is as follows:

D&d making a monster