Step #2 possibly (bit of trial and error here).
Lets assume that 00000000 = a space between words. T
Now we look for repeated blocks of numbersm (will assign ID tags to blocks)
while this does show some repeats, there are not enough (had to use ID tags A-Z, 1-9 ans a-k)
Of interest is that blocks of numbers seam to end in 6 or 7 and occasionally 8, suggesting a possible approach 'Table lookup' method, eg (Xcord, Ycord, Caps/lowercase), which would require searching for sub strings inside the number blocks and looking for repetition or patterns.
43454846 A
63656866 B
72635857 C
00000000
73757876 D
63646866 E
53545856 F
52748977 G
72637957 H
00000000
12138786 I
22237776 J
32336766 K
42435756 L
00000000
43454846 A
53553947 M
55566866 N
43454846 A
53553947 M
55566866 N
63645968 O
64656867 P
00000000
43454846 A
53555857 Q
63653947 R
73746866 S
72634967 T
00000000
12138786 I
22237776 J
32336766 K
42435756 L
00000000
33353836 U
43454847 V
53555856 W
63645957 X
72633958 Y
63725837 Z
00000000
43454846 A
53545857 1
63656866 B
00000000
33353836 U
53555856 W
62534847 2
42335957 3
52423958 4
42525837 5
0000000
53555857 Q
5254
00000000
12138786 I
22237776 J
32336766 K
42435756 L
00000000
63656866 B
53555856 W
73747877 6
72545957 7
62726968 8
72624958 9
00000000
63656867 a
53554846 b
43455857 c
72633947 d
00000000
73747877 6
63646866 E
53545857 1
74754967 e
64657968545
00000000
33353837 f
35364846 g
53555857 Q
55564645 h
63657877 i
737549387576
00000000
33357876 j
53555856 W
63653837 k
434537367375