I'm not convinced consultant style learning is THAT useful.
I know my AWS, I can be tossed into learning whatever product a client needs. The problem is that AWS is such a varied product that has its individual quirks that might not actually mesh properly with other services.
Sure I can probably deploy completely unfamiliar product fairly quicky, but without owning the product medium-longish term, you have no idea if it was actually properly deployed, meaning whatever you learned probably wasn't the best way to do a thing
Throwing something together based on unfamiliar components is indeed a bad move and likely to get you a bad reputation as a consultant. On the flip side, a permanent employee who wallows in analysis paralysis isn't adding any value to the company.
It all just depends on where the client culture fits on the "move fast and break things" spectrum.
I know my AWS, I can be tossed into learning whatever product a client needs. The problem is that AWS is such a varied product that has its individual quirks that might not actually mesh properly with other services.
Sure I can probably deploy completely unfamiliar product fairly quicky, but without owning the product medium-longish term, you have no idea if it was actually properly deployed, meaning whatever you learned probably wasn't the best way to do a thing