It seems the thought of Aegyptus out-producing Africa Proconsularis may be incorrect. We can find original ancient sources which indicate that it seems to have been the other way around, with North Africa providing 2/3 of Rome's grains. The reference providing this ratio, though lacking specific figures, can be found in Bellum Judaicum 2.383 by Josephus.
And besides the annual fruits of the earth, which maintain the
multitude of the Romans for eight months in the year...
Another ancient source, the Roman historian Tacitus, when discussing the volatility of Roman food supplies in Annals, 12.43, mentions Africa before Egypt when mentioning the sources of Roman grains, which can be taken as corroboration of Josephus:
in the past, Italy exported supplies for the legions into remote
provinces; nor is sterility the trouble now, but we cultivate Africa
and Egypt by preference, and the life of the Roman nation has been
staked upon cargo-boats and accidents.
The 1988 book The Economics of Agriculture on Roman Imperial Estates in North Africa
By Dennis P. Kehoe, lists these sources (footnote 8 on page 4) and tosses a couple of figures in (and points out the possible exaggeration on the part of Josephus):
These passages cannot be pressed too closely, but they do indicate
that Africa had become Rome's most important source of grain by the
reign of Nero. If Rome consumed 30-40 million modii of grain anually,
Africa would have supplied 20-27 million modii (130,000 - 180,000
metric tons).
Estimates based on the report of Egypt exporting 20 million modii, found in the fourth century work Epitome, if extrapolated out at the Josephus 1/3 ratio, would show a Roman consumption of 60 million modii, which has been argued as possibly too high. This discrepancy is mentioned in a wiki entry on the Cura Annonae:
Peter Garnsey combines the accounts of the author of the
fourth-century Epitome that 20 million modii of wheat came from Egypt
and Josephus' statement in the mid-first century AD that North Africa
provided twice the export of Egypt and that it supplied Rome eight
months of the year and Egypt supplied the other four, leaving a total
of 60 million modii imported to Rome. Garnsey finds this number too
high as this works out to 400,000 tons (800 million pounds) but only
200,000 tons was required for Augustus' first grain dole.4
The book Famine and Food Supply in the Graeco-Roman World: Responses to Risk and Crisis,(1989) by Peter Garnsey is the source referred in the above entry.
So, as we can see, entire books can and have been written on the subject of Roman grains.
Gleaning what information we can from historical sources from the first century, it seems that the North African region provided 2/3 of the grain consumed during the years of the early Roman Empire (the information may not reflect later times when Constantinople became an important part of the Empire) and can be summed up by the Kehoe statement from earlier:
Africa had become Rome's most important source of grain by the reign
of Nero.
(all above emphasis mine)
