Afaik there is no ceiling on the levels that can be achieved in year 2
Apart from the posters who have stated on various MN threads that they were told by their schools that no child would be awarded higher than a L3 in year 2. Or the teachers who have admitted they were pressured to keep levels down in case progress couldn't be maintained in KS2. Obviously I've no idea how frequently this has occurred, but as I tend to believe rather than disbelieve posters I do believe it has happened at least occasionally.
From level 3b now to a level 4c in May is a huge jump anyway and unlikely to be achieved. I cannot see what the problem is.
If a child has achieved a 3b and 3a by the first term of year 2 (so after 4 terms being taught the National Curriculum) then it is at least possible that they will sustain the accelerated progress they've already been demonstrating and succeed in achieving 1 to 2 further sublevels in another term.
Achieving L4 in year 2 remains exceptional, however many people post on this thread to attest to their own DCs results - see here.
There is no upper level that a child can attain in Year 2 neither are there pros or cons in attaining level 3 or level 6 in Y2 this year as the system will not exist next year
I would think the pro would be that the school has accurately recorded the child's level in year 2 and that there should be some system in place for transferring achievement under the old system to the new system. Even if there is no mechanism to manage this and every child has to be reassessed in year 3 as if from a blank slate
then there is still a benefit to accurate assessment as it should inform next steps, even in year 3. And that's without even mentioning that some schools are still sticking to levels for the time being.
I like the idea of being given a L3-5 task and then assessed accordingly. Would there be any cons than pros to this approach?
Well the 'con' with the L3-5 paper is you either achieve L4 or you don't - the paper's L4 is equivalent to a 4b, not a 4c. There is no banding of sublevels with the paper (which teacher assessment provides), therefore it depends on a child's performance on the day and whether the L4 questions are at least familiar to the child if they are 'only' a 4c.
Teacher assessment, on the other hand, allows the teacher to assess at L4c. So a child could be working at 4c (perhaps better to say they are working within L4 but are not secure at that level) and fail the paper but still be recorded at L4 via teacher assessment. The 'con' of teacher assessment is that it requires the teacher to have gathered the evidence over time and with maths in particular to provide appropriately differentiated work at L4.
Out of interest is it the current school that has assessed 3a/3b or the previous school? If the latter then you may find it a struggle as the current school won't have had as much time to gather their own evidence - in which case the 3-5 paper might be worth requesting.