Implement off-heap scoring for OSQ 4, 7, and 8 bit representations

[mccullocht]

Partial implementation of #15155

So far this is not any faster than the alternative. On an AMD RYZEN AI MAX+ 395

baseline:
Results:
recall  latency(ms)  netCPU  avgCpuCount     nDoc  topK  fanout  maxConn  beamWidth  quantized  visited  index(s)  index_docs/s  force_merge(s)  num_segments  index_size(MB)  vec_disk(MB)  vec_RAM(MB)  indexType
 0.913        1.635   1.630        0.997  1000000   100     100       32        250     8 bits     6824      0.00      Infinity            0.04             1         3759.67      3677.368      747.681       HNSW

candidate:
Results:
recall  latency(ms)  netCPU  avgCpuCount     nDoc  topK  fanout  maxConn  beamWidth  quantized  visited  index(s)  index_docs/s  force_merge(s)  num_segments  index_size(MB)  vec_disk(MB)  vec_RAM(MB)  indexType
 0.913        1.671   1.661        0.994  1000000   100     100       32        250     8 bits     6824      0.00      Infinity            0.04             1         3759.67      3677.368      747.681       HNSW

DO NOT MERGE
Performance observations: on an avx512 host the profiles are quite different. the original path spends a most time in dotProductBody512 followed by Int512Vector.reduceLanes(). the new path spends much more time in reduceLanes() but also spends more time loading from a memory segment for the input vectors -- a 128 bit load from a memory segment instead of a heap array. this could be memory latency but in that case why doesn't the load into the heap array show up in the profile?

[benwtrent]

@mccullocht maybe we only do the byte part of the comparisons off-heap? Then apply the corrections all on heap. I would assume applying corrections is pretty cheap, but even then if we did it in bulk, maybe on-heap bulk correction application is pretty fast?