There is no documentation to indicate that this level of throughput was REACHED with any regularity.
There were regularly published reports for on time performance, which was defined as reaching the terminal within 3 minutes of the the scheduled arrival time. If one examines those figures one can extrapolate a degree of reliability. These figures were generally well above 90% which makes moot any suggestion that the rush hour schedules were unreliable.
Admittedly, such indirect evidence is not as satisfying something more direct. However, that is all one has to go on for contemporary published data collected by the BOT/TA. Your skepticism of the reliability of those schedules, while healthy, is based on even less documentation.
One of the first things David Gunn did when he took over NYCT(A) in 1984 was to order that service be cut to levels that could reasonably be attained, rather than show on paper that more service than was actually operating was being scheduled.
There was a severe shortage of reliable operating equipment by 1984. Rush hour schedules had been reduced from their 1950's peak levels during the 1970's fiscal crisis. However, the TA did not have enough reliable operating equipment to operate even these reduced service levels. The TA's approach was twofold: to cancel certain trains on a regular basis and to permit an unacceptably large number of trains to break down en route. Mr. Gunn's "service cuts" were based on the availability of reliable equipment. These service cuts were not based on the question of whether or not such service levels could be reliably maintained had the equipment been available.