00:21:631 (21631|1) - i guess this LN's length can be exchanged with the another diff one? another diff is having a LN release but this finished the same time with 00:21:069 (21069|2,21069|0) - so I think better to exchange the length of LNs of 2 diffs.
00:51:444 (51444|4,51444|6) - should be triple too to be consistent with 00:52:194 (52194|2,52194|3,52194|5) - ?
i think 00:55:569 (55569|6,55569|4,55569|1,55756|2,55944|3,56131|4,56225|6,56319|0,56319|5,56319|4,56506|1,56694|2,56881|3,57069|6,57069|0,57069|4,57256|1,57444|2,57631|3,57631|5,57725|6,57819|1,57819|5,57819|4,58006|0,58194|6,58194|3,58194|1,58569|0,58569|2,58569|4) - the inverse gap or density should be bigger like 01:01:569 - 01:06:069? they're both basically repetitive in the music
01:17:994 (77994|1) - Would probs just move this down the 1/5 to the 1/4 at 01:17:975 - since there's not really anything in the music to suggest a grace.
01:50:225 (110225|1,110319|0,110319|1,110600|1,110694|0,110694|1,110975|1,111069|1,111069|0) - this is TOO tricky to hit, major diffspike imo