The choice between 5 great engineers and 1000 average engineers is, of course, an utterly false dilemma. No-one gets paid 200x average salary for being 'great'.

The choice between 5 great engineers and 1000 average engineers is, of course, an utterly false dilemma. No-one gets paid 200x average salary for being 'great'. The choice should be between hiring 5 great engineers PLUS 100 good engineers in support of these guys (and maybe 20 ancillary staff to answer the phones, and keep all the other distracting stuff out of the hair of the 55 engineering staff) vs 1000 average engineers. Maybe a few high-quality managers to make sure that 'greatness' is actually staying on schedule rather than just sitting around being really awesome, too. :-) Everyone has a different skill-set, and using your great engineers to do stuff outside their area of expertise is like using a oscilloscope as a hammer. This also leads to another corollary: that is, given 2 'great' engineers, if engineer A is a little less brilliant than engineer B, but can make effective use of a team of 20, while engineer B insists on doing everything 'his way' and can't work effectively with others, you're going to want engineer A.