76	b2-43 -- b3-9	1a	29	22	15		29
	b3-10 -- b3-16	1b			7	14
	b3-17-- b3-23	1c		7
	b3-24-- b3-35	2a	25	18	12		25
	b3-36-- b3-41	2b			6	13
	b3-42-- b4-1	2c		7
	b4-2-- b4-10	3a	22	15	9		22
	b4-11-- b4-16	3b			6	13
	b4-17-- b4-23	3c		7
20	b4-24-- b4-38	4a	29	20	15		29
	b4-39-- b4-43	4b			5	14
20	b4-44-- b5-1	4c		9
	b5-2-- b5-12	5a1	59	52	11		59
76	b5-13 -- b5-47	5a2			35
	b5-48 -- b5-53	5b			6	13
	b5-54 -- b6-3	5c		7
	b6-4 -- b6-18	6a	29	22	15		29
	b6-19 -- b6-25	6b			7	14
	b6-26-- b6-32	6c		7

This variant explores the effects if we try to use twice 20 as a way of defining limits for the glyph sequences. With red is marked the necessary changes.

Obviously these 20:s are not meant to be used in this way. We need therefore not divide 5a into two subsequences like this. Hb5-12 and Hb5-13 must belong together according to how these glyphs look. And Hb4-44 -- Hb4-45 are re-relocated to the wrong sequence, they should mark 'seating' in 4b.