I'm no expert on assembly-language programming, but probably 90% of the assembly I write on i386, amd64, RISC-V, and ARM is about 40 instructions: ldr, mov, bl, cmp, movs, push, pop, add, b/jmp, bl/blx/call, ret, str, beq/jz, bne/jnz, bhi/ja, bge/jge, cbz, stmia, ldmia, ldmdb, add/adds, addi, sub/subs, bx, xor/eor, and, or/orr, lsls/shl, lsrs/sar/shr, test/tst, inc, dec, lea, and slt, I think. Every once in a while you need a mul or a div or something. But the other 99% of the instruction set is either for optimized vectorized inner loops or for writing operating system kernels.
I think that the reason that i386 assembly (or amd64 assembly) is error-prone is something else, something it has in common with very simple architectures and instruction sets like that of the PDP-8.
What reason is that? (And, if it's not obvious, what are ARM/RISC-V doing that make them less bad?)
There are various minor details of one architecture or the other that make them more or less bug-prone, but those are minor compared to what they have in common.
None of this is because the instruction sets are complex. It would be closer to the mark to say that it's because they are simple.
I've not fleshed this out yet, but I think a relatively simple system would help deal with all the issues you mention in the first paragraph while allowing escape hatches.