Thorough Mathematical Analysis of This Situation is Difficult
What you are asking for is an evaluation of unknown "adjusted monsters" against unknown character classes at levels 1-7. We don't know what your GM is planning on doing, he could have a detailed excel spreadsheet where he accurately calculated a need to adjust the early game rules for an expanded standard deviation of expected character abilities, or he may view it as a non-issue and using a small lie to avoid arguing over a minor house rule.
The System Design is Built for Small Numeric Adjustments
Even with bounded accuracy we have to recall that a +1 on a d20 roll is still just a 5% increase. That isn't game changing. Given the system assumes the DM to be responsible to magic item distribution, which can have equitable or even greater impact on the numbers, it can be simply observed that the RAI actively discourage worrying about these things like getting to 20 in your primary stat early.
Stat Spread is Likely to Have a Far Greater Impact on the Game
However it is safe to assume that some, if not most, people WILL max out a stat due to the nature of the game, and in turn this will mean lowering some other stats. Slightly greater strengths, more pronounced weaknesses.
Look at some of the min/max arrays compared to the standard 15, 15, 15, 8, 8, 8:
16, 16, 11, 8, 8, 8
16, 16, 10, 9, 8, 8
16, 15, 13, 9, 8, 8
In pretty much every case what we are really doing is dropping a third stat. Moving from 3 strengths and 3 weaknesses to 2 strengths and 4 weaknesses. Other combinations are possible, but the general idea is that it is enhancing stat spread.
It is more likely for a party as a whole to have a glaring weakness, and (especially if feats are permitted) this is not going to be easily mitigated at later levels. Spending 4 extra points in point-buy to get 1 point stat means 4 less points are put into lower stats which mostly trade on a 1-for-1 basis. Level based stat gains are always a 1-for-1 basis so the long term cost of investing points into a high starting stats is substantial.
The rule adjustment's biggest impact is probably making the numbers more erratic. And considering failure state for PC's (typically death, or missing crucial information) is far less desirable than failure states for their opponents (Breaking News: Farmer Joe's prize cow returned by friendly adventurers! Sidebar: 800 year old local clan of Kobolds completely eradicated in 8 hour period) one could consider this to be slightly fudging the danger/difficulty upward rather than down.