Only await cache#get if it's a promise

[Eric-Tyrrell22]

This is admittedly a pretty small optimization ( maybe 5% ), but I figure I'd open a PR since I noticed it. I haven't fully tested this, but if there's interest in merging this I'll add some unit tests.

Ran the benchmark and got the following results:

Before

CacheableLookup#lookupAsync x 528,392 ops/sec ±0.52% (90 runs sampled)
CacheableLookup#lookupAsync.all x 532,076 ops/sec ±0.30% (89 runs sampled)
CacheableLookup#lookupAsync.all.ADDRCONFIG x 426,854 ops/sec ±0.41% (89 runs sampled)
CacheableLookup#lookup x 510,691 ops/sec ±0.61% (89 runs sampled)
CacheableLookup#lookup.all x 503,598 ops/sec ±0.61% (89 runs sampled)
CacheableLookup#lookup.all.ADDRCONFIG x 420,734 ops/sec ±0.47% (90 runs sampled)
dns#lookup x 2,782 ops/sec ±3.95% (67 runs sampled)
dns#lookup.all x 3,182 ops/sec ±6.14% (75 runs sampled)
dns#lookup.all.ADDRCONFIG x 2,448 ops/sec ±2.92% (69 runs sampled)

After

CacheableLookup#lookupAsync x 552,957 ops/sec ±0.62% (86 runs sampled)
CacheableLookup#lookupAsync.all x 556,641 ops/sec ±0.50% (89 runs sampled)
CacheableLookup#lookupAsync.all.ADDRCONFIG x 438,904 ops/sec ±0.46% (89 runs sampled)
CacheableLookup#lookup x 533,336 ops/sec ±0.54% (86 runs sampled)
CacheableLookup#lookup.all x 533,434 ops/sec ±0.57% (85 runs sampled)
CacheableLookup#lookup.all.ADDRCONFIG x 432,496 ops/sec ±0.51% (88 runs sampled)
dns#lookup x 2,378 ops/sec ±3.45% (69 runs sampled)
dns#lookup.all x 2,834 ops/sec ±5.86% (74 runs sampled)
dns#lookup.all.ADDRCONFIG x 2,656 ops/sec ±5.46% (65 runs sampled)