SEN. PETE DOMENICI, chairman of the Senate's Budget Committee, delivered a gentle but unambiguous caution the other day to the Pentagon. There's going to be a large increase in real defense spending, he said. But it probably won't be as large as the figures currently projected for the years ahead in the budget documents.

If the budget is to move toward balance, further rounds of severe cuts in spending will be required. In that process, Mr. Domenici observed, very little if anything is going to be sacred, and the military budget is going to get the same scrutiny as all the rest. Some procurement schedules, and (the introduction of new defense systems, may have to be spread out over longer periods than the Defense Department currently expects.

The defense estimates in President Reagan's budget, particularly for the years after 1982, were not calculated from specific program requirements. They were calcualted, in great haste, by a very new admnistration, as signals to indicate a degree of urgency and a general intention. Many people agree that defense spending had fallen to a point by the middle 1970s that required an accelerated rate of rebuilding. But there are also limits to the speed at which money can be spent efficiently.

Government spending responds to the unpredictable movements of the national economy, and budgeting is necessarily a constant process of revision. Political debate gives an artificial and misleading air of precision to a budget, as partisans insist on their own forecasts and write specific numbers into law. Mr. Domenici was making the valuable point -- especially valuable, in this year of extraordinary budget debate -- that all of these estimates and projections have to be used with the knowledge that reality can rapidly seize and bend them. A budget sets policy, but it can't promise to control the future. With massive strains already becoming apparent in the next several years' budgeting, it seems pretty clear that not all of the defense figures can be fitted in.

What will consitute success for the Reagan budgets? As the inevitable question goes, is the strategy going to work? The economic indicators would not have to follow the projections precisely, to persuade people that the administration's budget program is having a healthy effect. The Reagan plan promises very high economic growth rate beginning next year. Suppose that the growth rate is only half that high, but that unemployment declines and both inflation and interest fall sharply lower. Wouldn't most people accept that as success? The test for defense will be similar. Except perhaps for defense contractors, few will greatly care whether military spending touches a certain published figure on the last day of fiscal year 1984. The real question will be whether the administration can show that it has built a defense capability stronger and more reliable than today's.