Safety of an operation is dependent on multiple variables, of which crew competency is a factor. Routinely airlines will operate with a zero experience pilot in line training, with or without a safety pilot dependent on the local rules. That is not going to change, at least in this universe, people start at the beginning and add experience.
Mitigation of the risks of low experience or assumed limited competency comes from increased oversight, more restrictive operational limits, and adding good equipment and hopefully some balance of limited experience with pilots that have a modicum of experience but more importantly competency. Irrespective of how well or inadequate the original training and experience was, the majority of the operational learning will occur on line, where there is experiential reinforcement of hopefully remembered concepts and rules.
For basic flying required to fly a hi capacity RPT aircraft in a mature airline, you can consider the basics that exist out of the military training program, and remove the tactical stuff in the most part. Out of a 230-250 hr flight training program, around 150 hrs of that is related to "flight training", the rest is focused towards military centric stuff, formation, low flying, aerobatics, and initial weapons training, which add to the total competency, but are focused outside of the most basic civil task. Out of 150 hrs, the first 20 odd hours are gaining basic flight skills, getting around in a semblance of order from TO to LDG. The rest is related to learning instrument flying skills, navigation, and becoming comfortable in night operations. We are accepting of a minuscule amount of time in learning to fly multi engine aircraft, and in all honesty, for the most part that time is wasted, the concepts and skills of ME are pretty basic. As an industry since inception we have accepted new pilots being turned around immediately to train new pilots, was ever thus. In the main, these instructors do a creditable job, but it is inherently a lousy concept, but it is what we have always accepted as a norm. The use of simulators has the potential to be a better way of learning, and also it may result in considerable lack of confidence in the student, which can come back and bite. Incorporating basic flight training to beyond solo, simulator, and some advanced flying to give confidence would be a reasonable way to get a pilot into a program with multi crew and oversight that can manage the low experience. That is pretty close to the concept of the MPL.
Centaurus' concern of the new low experience pilot sitting solo in the front end as the old guy, like Centaurus and me have croaked defines the basic level of competency that needs to be assured to have the low time pilot sit in the aircraft without any other assistance.
I believe we need training that speaks more to the actual task the pilot will undertake, appropriate level of supervision, and associated operational limitations that maintain safety of operation. That is more or less the same issue that has existed since 1903, WWI, WW2, 10,000 day war, etc... the only difference is that we continue to improve the training tools that exist, but as always, effective utilisation of the tools lags technology as the rules move at glacial speed. Personally, I would not take issue to competent instructors teaching the majority of sim training in FBT (for nav and procedures including IF), and using FFS for limited tasks related to pattern work, unusual attitudes and similar tasks. Decision making, CRM etc, and much NNCL training is suited to PTT/FBT. For the aircraft flying itself, fly one of the modern LSA's, they outperform most "complex" aircraft and are cost effective, or stay with a 150, or better yet, put the J3 back into production. For advanced flying, strap on an S2A/C or similar. For twin training, it is a part of the FFS, and would give more rational training safer than we do today in a PA30/34 and similar.
There is not a lack of skills or training, there is a lack of imagination and effective resource use.