Cargo Dragon for crew?

Using a cargo Dragon for manned crew to ISS.

First, Launch Abort.

One issue is the super dracos won't fit in the current cargo version of Dragon. But... they probably will in soon to be built cargo dragons- they're required for landing on land.

So, a current cargo Dragon, such as the one on the pad now, won't have a launch abort system, and flying crew without a LAS would be utter madness that NASA would never go along with (except they did for Shuttle, which had no LAS).
Shuttle flew its first launch maned. Was there a LAS? Well yes, for the first four flights; SR-71 ejection seats. Columbia had them, as did the glide test orbiter Enterprise. But, for Columbia, there was a tiny drawback; the seats could indeed allow the crew to eject at any point between launch and 100,000 feet (so prior to SRB burnout. This made those seats a true LAS system. The slight drawback I mentioned is that using the seats at *any* time between liftoff and seat max altitude would be fatal for the crew (the IR from the SRBs would be well above what the suits could stand; the ejection profile would take the crew too close to the SRB plumes). The only time the seats would be survivable would be for a pad fire (The crew would have about a 50% chance of landing alive so they could be killed on the ground by the time the vehicle's explosion, due to having landed in their chutes just yards from the pad) or for a post-reentry low supersonic to subsonic ejection during controlled flight. (in this latter case, and ONLY in this latter case, could the seats be used with a reasonable chance for survival (of more than a few seconds).

What this means is that you could put a teddy bear in a crewed cargo dragon, name it "LAS", and it'd be every bit as effective as the ejection seat LAS on Colombia's first four launches.

But, to fully test the Dragon's super Draco LAS system, NASA has decreed that you have to do it twice; a pad abort (that one is okay) and a fully up launch vehicle taking the Dragon through MaxQ for an abort test. But, NASA has always insisted on this, right? Let's look back at history...

Mercury? They did test the LAS, but with a suborbital Mercury-Redstone booster. The test was, well, interesting in how it worked out (not at all as planned). The rocket shut down just inches from the pad, and dropped back. Then, the LAS tower ejected (leaving the capsule behind) and reached 4000 feet. That left the capsule on the fueled, damaged, booster. The capsule then triggered its parachutes. It was a mess. They never flight-tested the Mercury LAS again.

Gemini? It had no actual LAS, it used ejection seats.

Apollo? They used a "Little Joe", a small booster, to test the system 2.9 seconds into the flight, NOT at Max-Q with a Saturn launch vehicle. This remains the only successful flight test, or use, of an American LAS.

But, NASA is requiring SpaceX to use a full-up F9 1.1 (including a real second stage that won't be used at all!) for the Max-Q (maximum dynamic pressure) LAS test (about 57 seconds into the flight).  

Could cargo Dragon, as is, carry a crew to ISS? Pretty much, yes. Must said, after the first Dragon mission to ISS, that a stow-away would have survived. Dragon already has thermal control abilities (as evidenced by it's carraige of freezers and refrigerators for experiments). You don't actually need a flight control system (just use it as is). What you do need is life support, but the hard parts are already there; thermal control and pressurization. All you really need is oxygen and CO2 scrubbing. An O2 tank , just strapped to a bulkhead and with an adjustable bleed valve would handle the former. Lithium hydroxide canisters with a computer CPU fan for air circulation would handle the latter.
Communications? Dragon already has them; you'd need microphones to attach to the control system jacks. That's it.

All that actually remains is the seats.

So, how do you put seats in a cargo Dragon? It already has the bulkhead attach points. You just need the seats (acceleration couches). Canvas slung between aluminum tubing frames has been used before (Such as on Apollo). But, the needed seats don't currently exist. So, they'd need to be built. SpaceX could surly do this, but if not, it could be done by a couple of high school students with access to a high school metal shop, plus a sewing machine.  It might take a couple of days. Or, have NASA do it, for a billion and a half dollars, and they might have it ready in a year and a half.

The actual tricky part in using Dragon as it exists isn't getting crew to orbit, but getting the crew into the Dragon on the pad. There's no easy way up, and no way of fast exit in case of a pad emergency. The latter could be remedies by long zip lines (commercially available). Getting the crew up there and in in the first place is much harder. The fastest solution (could be done today) is the Falcon9's normal launch prep; it's raised on the strongback a few hours before launch, so put the crew in prior to raising.

So, if we were serous, how long would it take to prep a cargo dragon to take a crew of two or three (thus replacing Soyuz - a crew of seven is harder) to ISS, as safely as a Shuttle? A few weeks at most. The biggest obstacle isn't technical, but bureaucratic.