RE: Cost-effectivenesss of highly-accurate clocks for NTP