Major - P3 It seems there was a change in the way column alignment is performed such that columns tend to become more severely misaligned.
Below is an example where particularly after resident mem the columns expand and contract greatly, making the output more difficult to follow.
*0 3768 1988 *0 112 662|0 3.6 7.1 0 9.8G 5.5G 306|162 398|222 2m 11m 3732 mmsconfig1 PRI 02:34:53
4 3670 1850 *0 60 643|0 3.7 7.2 0 9.8G 5.5G 131|83 335|227 1m 11m 3734 mmsconfig1 PRI 02:34:58
*0 3415 1906 *0 46 609|0 3.8 7.2 0 9.8G 5.5G 357|183 374|185 1m 9m 3734 mmsconfig1 PRI 02:35:03
2 3780 1868 *0 72 597|0 3.9 7.2 0 9.8G 5.5G 190|136 260|170 2m 10m 3735 mmsconfig1 PRI 02:35:08
3 3457 1906 *0 44 584|0 4.0 7.2 0 9.8G 5.6G 128|72 245|140 1m 9m 3739 mmsconfig1 PRI 02:35:13
insert query update delete getmore command % dirty % used flushes vsize res qr|qw ar|aw netIn netOut conn set repl time
*0 3496 1387 *0 33 653|0 4.0 7.2 0 9.8G 5.6G 84|17 586|232 1m 10m 3763 mmsconfig1 PRI 02:35:18
*0 3623 1630 *0 40 612|0 4.1 7.2 0 10.0G 5.6G 52|25 579|233 1m 11m 3846 mmsconfig1 PRI 02:35:23
4 3864 1794 *0 82 618|0 0.7 7.2 1 10.1G 5.7G 123|61 384|214 1m 12m 3914 mmsconfig1 PRI 02:35:28
1 3427 1961 *0 82 598|0 2.0 7.3 0 10.2G 5.7G 90|55 242|169 2m 10m 3918 mmsconfig1 PRI 02:35:33
2 3314 1708 *0 59 554|0 2.5 7.3 0 10.2G 5.7G 55|32 246|158 1m 10m 3919 mmsconfig1 PRI 02:35:38
3 3336 1701 *0 59 590|0 2.9 7.3 0 10.2G 5.8G 154|87 272|172 1m 10m 3919 mmsconfig1 PRI 02:35:43
1 3308 1768 *0 98 647|0 3.2 7.3 0 10.2G 5.8G 159|76 265|149 1m 10m 3920 mmsconfig1 PRI 02:35:49
2 3360 1696 *0 62 621|0 3.5 7.3 0 10.2G 5.8G 51|35 297|161 1m 9m 3920 mmsconfig1 PRI 02:35:54
6 3507 1754 *0 44 606|0 3.6 7.4 0 10.2G 5.8G 344|116 388|237 1m 9m 3920 mmsconfig1 PRI 02:35:59
1 3486 1856 *0 70 595|0 3.8 7.4 0 10.2G 5.8G 41|34 360|260 2m 10m 3922 mmsconfig1 PRI 02:36:04
insert query update delete getmore command % dirty % used flushes vsize res qr|qw ar|aw netIn netOut conn set repl time
3 3319 1839 *0 56 585|0 3.9 7.4 0 10.3G 5.8G 211|140 307|249 1m 10m 3938 mmsconfig1 PRI 02:36:09
3 3102 1687 *0 53 613|0 4.0 7.4 0 10.3G 5.8G 28|62 476|353 1m 9m 3963 mmsconfig1 PRI 02:36:14
1 3520 1666 *0 42 547|0 4.1 7.4 0 10.3G 5.9G 381|166 401|182 1m 9m 3963 mmsconfig1 PRI 02:36:19
1 3682 1515 *0 59 603|0 4.2 7.4 0 10.4G 5.9G 364|181 430|224 1m 10m 3963 mmsconfig1 PRI 02:36:24
4 3519 1650 *0 63 632|0 4.0 7.4 1 10.4G 5.9G 357|184 487|238 1m 10m 3963 mmsconfig1 PRI 02:36:29
Off the top of my head I believe the behavior used to be to continue expanding right as needed but never contracting back left. So at worst the width expands to fit the widest ever observed such that from then on the alignment's always exact, even if values later "contract" from their max width.