Yes, I think that works best. You need an engineering leader who has some understanding of all of the important problem areas the company faces, even though he/she is unlikely to be the company's leading expert in any of them, or more than one of them, so good decisions can be made.