Yes it can go both ways. Just as natural disasters can cause economic growth, so can wars.
The main reason: economics for business does not count the cost of externialities, and does not count the costs of "bads". For instance, for pollution, such as oil spill, the loss of sales may be counted, but so is the jobs, wages, and services created to clean up the mess is actually a good, as incomes and the GDP are increased. The final tally of the loss of sales versus the increased earning from the spill are all the count. Externialities such as environmental damage, etc is not considered.
During war, much is the same as an oil spill.
There are economic losses that occur. Loss of production facilities, transactional services and trading systems, less production due to loss of personnell etc. However there are economic gains: new productions of equipment, facilities, mechanical innovations to improve efficiency, etc. Some businesses thrived during wartime, some floundered.
A tally would need to be done for during wars to see the trending.
There would be three concerns as well: public sector agencies would be the main purchasing agent for all efforts, creating long term debt. The transition of military to general free market economy would itself have costs, and centralized economies have failed to provide wealth or stability for long periods of time.