The instruction set is actually quite tiny.

To try to keep many instructions 8-bit large, a few bits are used to select the instruction, to leave as many bits as possible to specify what the instruction should operate on. That is also a reason why an old-timer can learn to program a chip directly in machine code - much of the total combinations possible can be remembered using a few rules.

The small instruction set can be seen as a limitation, but gives the advantage that most instructions are tiny, which is good for a chip designed when the code space (ROM, EPROM, ...) was very expensive. The tiny instructions makes it trivial to decode the instructions, which is another reason for the success of the 805x cores. It delivers a "you know what you get". On the other hand, the C compiler vendors have to invest much blood to get good code generation.

Edit: Fixed spelling in subject line